{ "cells": [ { "cell_type": "markdown", "metadata": { "hide_input": true }, "source": [ "## The data block API" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [], "source": [ "from fastai.gen_doc.nbdoc import *\n", "from fastai.basics import *\n", "np.random.seed(42)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The data block API lets you customize the creation of a [`DataBunch`](/basic_data.html#DataBunch) by isolating the underlying parts of that process in separate blocks, mainly:\n", " 1. Where are the inputs and how to create them?\n", " 1. How to split the data into a training and validation sets?\n", " 1. How to label the inputs?\n", " 1. What transforms to apply?\n", " 1. How to add a test set?\n", " 1. How to wrap in dataloaders and create the [`DataBunch`](/basic_data.html#DataBunch)?\n", " \n", "Each of these may be addressed with a specific block designed for your unique setup. Your inputs might be in a folder, a csv file, or a dataframe. You may want to split them randomly, by certain indices or depending on the folder they are in. You can have your labels in your csv file or your dataframe, but it may come from folders or a specific function of the input. You may choose to add data augmentation or not. A test set is optional too. Finally you have to set the arguments to put the data together in a [`DataBunch`](/basic_data.html#DataBunch) (batch size, collate function...)\n", "\n", "The data block API is called as such because you can mix and match each one of those blocks with the others, allowing for a total flexibility to create your customized [`DataBunch`](/basic_data.html#DataBunch) for training, validation and testing. The factory methods of the various [`DataBunch`](/basic_data.html#DataBunch) are great for beginners but you can't always make your data fit in the tracks they require.\n", "\n", "\"Mix\n", "\n", "As usual, we'll begin with end-to-end examples, then switch to the details of each of those parts." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Examples of use" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's begin with our traditional MNIST example." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_TINY)\n", "tfms = get_transforms(do_flip=False)\n", "path.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train/3'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train/7')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "(path/'train').ls()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In [`vision.data`](/vision.data.html#vision.data), we can create a [`DataBunch`](/basic_data.html#DataBunch) suitable for image classification by simply typing:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = ImageDataBunch.from_folder(path, ds_tfms=tfms, size=64)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This is a shortcut method which is aimed at data that is in folders following an ImageNet style, with the [`train`](/train.html#train) and `valid` directories, each containing one subdirectory per class, where all the labelled pictures are. There is also a `test` directory containing unlabelled pictures. \n", "\n", "Here is the same code, but this time using the data block API, which can work with any style of a dataset. All the stages, which will be explained below, can be grouped together like this:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = (ImageList.from_folder(path) #Where to find the data? -> in path and its subfolders\n", " .split_by_folder() #How to split in train/valid? -> use the folders\n", " .label_from_folder() #How to label? -> depending on the folder of the filenames\n", " .add_test_folder() #Optionally add a test set (here default name is test)\n", " .transform(tfms, size=64) #Data augmentation? -> use tfms with a size of 64\n", " .databunch()) #Finally? -> use the defaults for conversion to ImageDataBunch" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we can look at the created DataBunch:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "data.show_batch(3, figsize=(6,6), hide_axis=False)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's look at another example from [`vision.data`](/vision.data.html#vision.data) with the planet dataset. This time, it's a multiclassification problem with the labels in a csv file and no given split between valid and train data, so we use a random split. The factory method is:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "planet = untar_data(URLs.PLANET_TINY)\n", "planet_tfms = get_transforms(flip_vert=True, max_lighting=0.1, max_zoom=1.05, max_warp=0.)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
image_nametags
0train_31112clear primary
1train_4300partly_cloudy primary water
2train_39539clear primary water
3train_12498agriculture clear primary road
4train_9320clear primary
\n", "
" ], "text/plain": [ " image_name tags\n", "0 train_31112 clear primary\n", "1 train_4300 partly_cloudy primary water\n", "2 train_39539 clear primary water\n", "3 train_12498 agriculture clear primary road\n", "4 train_9320 clear primary" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.read_csv(planet/\"labels.csv\").head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = ImageDataBunch.from_csv(planet, folder='train', size=128, suffix='.jpg', label_delim = ' ', ds_tfms=planet_tfms)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With the data block API we can rewrite this like that:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/planet_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/planet_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/planet_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/planet_tiny/models')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "planet.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
image_nametags
0train_31112clear primary
1train_4300partly_cloudy primary water
2train_39539clear primary water
3train_12498agriculture clear primary road
4train_9320clear primary
\n", "
" ], "text/plain": [ " image_name tags\n", "0 train_31112 clear primary\n", "1 train_4300 partly_cloudy primary water\n", "2 train_39539 clear primary water\n", "3 train_12498 agriculture clear primary road\n", "4 train_9320 clear primary" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.read_csv(planet/\"labels.csv\").head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = (ImageList.from_csv(planet, 'labels.csv', folder='train', suffix='.jpg')\n", " #Where to find the data? -> in planet 'train' folder\n", " .split_by_rand_pct()\n", " #How to split in train/valid? -> randomly with the default 20% in valid\n", " .label_from_df(label_delim=' ')\n", " #How to label? -> use the second column of the csv file and split the tags by ' '\n", " .transform(planet_tfms, size=128)\n", " #Data augmentation? -> use tfms with a size of 128\n", " .databunch()) \n", " #Finally -> use the defaults for conversion to databunch" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "data.show_batch(rows=2, figsize=(9,7))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The data block API also allows you to get your data together in problems for which there is no direct [`ImageDataBunch`](/vision.data.html#ImageDataBunch) factory method. For a segmentation task, for instance, we can use it to quickly get a [`DataBunch`](/basic_data.html#DataBunch). Let's take the example of the [camvid dataset](http://mi.eng.cam.ac.uk/research/projects/VideoRec/CamVid/). The images are in an 'images' folder and their corresponding mask is in a 'labels' folder." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "camvid = untar_data(URLs.CAMVID_TINY)\n", "path_lbl = camvid/'labels'\n", "path_img = camvid/'images'" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We have a file that gives us the names of the classes (what each code inside the masks corresponds to: a pedestrian, a tree, a road...)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array(['Animal', 'Archway', 'Bicyclist', 'Bridge', 'Building', 'Car', 'CartLuggagePram', 'Child', 'Column_Pole',\n", " 'Fence', 'LaneMkgsDriv', 'LaneMkgsNonDriv', 'Misc_Text', 'MotorcycleScooter', 'OtherMoving', 'ParkingBlock',\n", " 'Pedestrian', 'Road', 'RoadShoulder', 'Sidewalk', 'SignSymbol', 'Sky', 'SUVPickupTruck', 'TrafficCone',\n", " 'TrafficLight', 'Train', 'Tree', 'Truck_Bus', 'Tunnel', 'VegetationMisc', 'Void', 'Wall'], dtype=' in path_img and its subfolders\n", " .split_by_rand_pct()\n", " #How to split in train/valid? -> randomly with the default 20% in valid\n", " .label_from_func(get_y_fn, classes=codes)\n", " #How to label? -> use the label function on the file name of the data\n", " .transform(get_transforms(), tfm_y=True, size=128)\n", " #Data augmentation? -> use tfms with a size of 128, also transform the label images\n", " .databunch())\n", " #Finally -> use the defaults for conversion to databunch" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "data.show_batch(rows=2, figsize=(7,5))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Another example for object detection. We use our tiny sample of the [COCO dataset](http://cocodataset.org/#home) here. There is a helper function in the library that reads the annotation file and returns the list of images names with the list of labelled bboxes associated to it. We convert it to a dictionary that maps image names with their bboxes and then write the function that will give us the target for each image filename." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "coco = untar_data(URLs.COCO_TINY)\n", "images, lbl_bbox = get_annotations(coco/'train.json')\n", "img2bbox = dict(zip(images, lbl_bbox))\n", "get_y_func = lambda o:img2bbox[o.name]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The following code is very similar to what we saw before. The only new addition is the use of a special function to collate the samples in batches. This comes from the fact that our images may have multiple bounding boxes, so we need to pad them to the largest number of bounding boxes." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = (ObjectItemList.from_folder(coco)\n", " #Where are the images? -> in coco and its subfolders\n", " .split_by_rand_pct() \n", " #How to split in train/valid? -> randomly with the default 20% in valid\n", " .label_from_func(get_y_func)\n", " #How to find the labels? -> use get_y_func on the file name of the data\n", " .transform(get_transforms(), tfm_y=True)\n", " #Data augmentation? -> Standard transforms; also transform the label images\n", " .databunch(bs=16, collate_fn=bb_pad_collate)) \n", " #Finally we convert to a DataBunch, use a batch size of 16,\n", " # and we use bb_pad_collate to collate the data into a mini-batch" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "data.show_batch(rows=2, ds_type=DatasetType.Valid, figsize=(6,6))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "But vision isn't the only application where the data block API works. It can also be used for text and tabular data. With our sample of the IMDB dataset (labelled texts in a csv file), here is how to get the data together for a language model." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.text import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "imdb = untar_data(URLs.IMDB_SAMPLE)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data_lm = (TextList\n", " .from_csv(imdb, 'texts.csv', cols='text')\n", " #Where are the text? Column 'text' of texts.csv\n", " .split_by_rand_pct()\n", " #How to split it? Randomly with the default 20% in valid\n", " .label_for_lm()\n", " #Label it for a language model\n", " .databunch())\n", " #Finally we convert to a DataBunch" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
idxtext
0! ! ! xxmaj finally this was directed by the guy who did xxmaj big xxmaj xxunk ? xxmaj must be a replay of xxmaj jonestown - hollywood style . xxmaj xxunk ! xxbos xxmaj this is a extremely well - made film . xxmaj the acting , script and camera - work are all first - rate . xxmaj the music is good , too , though it is
1, co - billed with xxup the xxup xxunk xxup vampire . a xxmaj spanish - xxmaj italian co - production where a series of women in a village are being murdered around the same time a local count named xxmaj yanos xxmaj xxunk is seen on xxunk , riding off with his ' man - eating ' dog behind him . \\n \\n xxmaj the xxunk already suspect
2sad relic that is well worth seeing . xxbos i caught this on the dish last night . i liked the movie . i xxunk to xxmaj russia 3 different times ( xxunk our 2 kids ) . i ca n't put my finger on exactly why i liked this movie other than seeing \" bad \" turn \" good \" and \" good \" turn \" semi - bad
3pushed him along . xxmaj the story ( if it can be called that ) is so full of holes it 's almost funny , xxmaj it never really explains why the hell he survived in the first place , or needs human flesh in order to survive . xxmaj the script is poorly written and the dialogue xxunk on just plane stupid . xxmaj the climax to movie (
4the xxunk of the xxmaj xxunk xxmaj race and had the xxunk of some of those racist xxunk . xxmaj fortunately , nothing happened like the incident in the movie where the young xxmaj caucasian man went off and started shooting at a xxunk gathering . \\n \\n i can only hope and pray that nothing like that ever will happen . \\n \\n xxmaj so is \"
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "data_lm.show_batch()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For a classification problem, we just have to change the way labeling is done. Here we use the csv column `label`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data_clas = (TextList.from_csv(imdb, 'texts.csv', cols='text')\n", " .split_from_df(col='is_valid')\n", " .label_from_df(cols='label')\n", " .databunch())" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
texttarget
xxbos xxmaj raising xxmaj victor xxmaj vargas : a xxmaj review \\n \\n xxmaj you know , xxmaj raising xxmaj victor xxmaj vargas is like sticking your hands into a big , xxunk bowl of xxunk . xxmaj it 's warm and gooey , but you 're not sure if it feels right . xxmaj try as i might , no matter how warm and gooey xxmaj raising xxmajnegative
xxbos xxup the xxup shop xxup around xxup the xxup corner is one of the xxunk and most feel - good romantic comedies ever made . xxmaj there 's just no getting around that , and it 's hard to actually put one 's feeling for this film into words . xxmaj it 's not one of those films that tries too hard , nor does it come up withpositive
xxbos xxmaj now that xxmaj che(2008 ) has finished its relatively short xxmaj australian cinema run ( extremely limited xxunk screen in xxmaj xxunk , after xxunk ) , i can xxunk join both xxunk of \" xxmaj at xxmaj the xxmaj movies \" in taking xxmaj steven xxmaj soderbergh to task . \\n \\n xxmaj it 's usually satisfying to watch a film director change his style /negative
xxbos xxmaj this film sat on my xxmaj xxunk for weeks before i watched it . i xxunk a self - indulgent xxunk flick about relationships gone bad . i was wrong ; this was an xxunk xxunk into the screwed - up xxunk of xxmaj new xxmaj xxunk . \\n \\n xxmaj the format is the same as xxmaj max xxmaj xxunk ' \" xxmaj la xxmaj xxunkpositive
xxbos xxmaj many neglect that this is n't just a classic due to the fact that it 's the first xxup 3d game , or even the first xxunk - up . xxmaj it 's also one of the first xxunk games , one of the xxunk definitely the first ) truly claustrophobic games , and just a pretty well - xxunk gaming experience in general . xxmaj with graphicspositive
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "data_clas.show_batch()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Lastly, for tabular data, we just have to pass the name of our categorical and continuous variables as an extra argument. We also add some [`PreProcessor`](/data_block.html#PreProcessor)s that are going to be applied to our data once the splitting and labelling is done." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.tabular import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "adult = untar_data(URLs.ADULT_SAMPLE)\n", "df = pd.read_csv(adult/'adult.csv')\n", "dep_var = 'salary'\n", "cat_names = ['workclass', 'education', 'marital-status', 'occupation', 'relationship', 'race', 'sex', 'native-country']\n", "cont_names = ['education-num', 'hours-per-week', 'age', 'capital-loss', 'fnlwgt', 'capital-gain']\n", "procs = [FillMissing, Categorify, Normalize]" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "data = (TabularList.from_df(df, path=adult, cat_names=cat_names, cont_names=cont_names, procs=procs)\n", " .split_by_idx(valid_idx=range(800,1000))\n", " .label_from_df(cols=dep_var)\n", " .databunch())" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
workclasseducationmarital-statusoccupationrelationshipracesexnative-countryeducation-num_naeducation-numhours-per-weekagecapital-lossfnlwgtcapital-gaintarget
?DoctorateMarried-civ-spouse?HusbandAmer-Indian-EskimoMaleUnited-StatesFalse2.3157-0.03561.7161-0.2164-1.1496-0.1459>=50k
PrivateSome-collegeNever-marriedSalesOwn-childWhiteMaleUnited-StatesFalse-0.0312-0.4406-1.4357-0.2164-0.1893-0.1459<50k
PrivateSome-collegeNever-marriedProtective-servOwn-childWhiteMaleUnited-StatesFalse-0.0312-2.0606-1.2891-0.21641.1154-0.1459<50k
PrivateHS-gradMarried-civ-spouseHandlers-cleanersWifeWhiteFemaleMexicoFalse-0.4224-0.0356-0.7027-0.21640.0779-0.1459>=50k
PrivateHS-gradMarried-civ-spouseTech-supportHusbandWhiteMaleUnited-StatesFalse-0.42243.2043-0.1163-0.2164-0.6858-0.1459>=50k
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "data.show_batch()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 1: Provide inputs" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The basic class to get your inputs into is the following one. It's also the same class that will contain all of your labels (hence the name [`ItemList`](/data_block.html#ItemList))." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class ItemList[source][test]

\n", "\n", "> ItemList(**`items`**:`Iterator`\\[`T_co`\\], **`path`**:`PathOrStr`=***`'.'`***, **`label_cls`**:`Callable`=***`None`***, **`inner_df`**:`Any`=***`None`***, **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **`x`**:`ItemList`=***`None`***, **`ignore_empty`**:`bool`=***`False`***)\n", "\n", "
×

Tests found for ItemList:

Some other tests where ItemList is used:

  • pytest -sv tests/test_data_block.py::test_category [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_filter_by_folder [source]
  • pytest -sv tests/test_data_block.py::test_multi_category [source]
  • pytest -sv tests/test_data_block.py::test_regression [source]
  • pytest -sv tests/test_data_block.py::test_split_subsets [source]
  • pytest -sv tests/test_data_block.py::test_splitdata_datasets [source]

To run tests please refer to this guide.

\n", "\n", "A collection of items with `__len__` and `__getitem__` with `ndarray` indexing semantics. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList, title_level=3)" ] }, { "cell_type": "markdown", "metadata": { "hide_input": true }, "source": [ "This class regroups the inputs for our model in `items` and saves a `path` attribute which is where it will look for any files (image files, csv file with labels...). `label_cls` will be called to create the labels from the result of the label function, `inner_df` is an underlying dataframe, and `processor` is to be applied to the inputs after the splitting and labeling." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It has multiple subclasses depending on the type of data you're handling. Here is a quick list:\n", " - [`CategoryList`](/data_block.html#CategoryList) for labels in classification\n", " - [`MultiCategoryList`](/data_block.html#MultiCategoryList) for labels in a multi classification problem\n", " - [`FloatList`](/data_block.html#FloatList) for float labels in a regression problem\n", " - [`ImageList`](/vision.data.html#ImageList) for data that are images\n", " - [`SegmentationItemList`](/vision.data.html#SegmentationItemList) like [`ImageList`](/vision.data.html#ImageList) but will default labels to [`SegmentationLabelList`](/vision.data.html#SegmentationLabelList)\n", " - [`SegmentationLabelList`](/vision.data.html#SegmentationLabelList) for segmentation masks\n", " - [`ObjectItemList`](/vision.data.html#ObjectItemList) like [`ImageList`](/vision.data.html#ImageList) but will default labels to `ObjectLabelList`\n", " - `ObjectLabelList` for object detection\n", " - [`PointsItemList`](/vision.data.html#PointsItemList) for points (of the type [`ImagePoints`](/vision.image.html#ImagePoints))\n", " - [`ImageImageList`](/vision.data.html#ImageImageList) for image to image tasks\n", " - [`TextList`](/text.data.html#TextList) for text data\n", " - [`TextList`](/text.data.html#TextList) for text data stored in files\n", " - [`TabularList`](/tabular.data.html#TabularList) for tabular data\n", " - [`CollabList`](/collab.html#CollabList) for collaborative filtering" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can get a little glimpse of how [`ItemList`](/data_block.html#ItemList)'s basic attributes and methods behave with the following code examples." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (3 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/labels.csv,/home/ubuntu/.fastai/data/mnist_tiny/history.csv,/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from fastai.vision import *\n", "path_data = untar_data(URLs.MNIST_TINY)\n", "il_data = ItemList.from_folder(path_data, extensions=['.csv'])\n", "il_data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Here is how to access the path of [`ItemList`](/data_block.html#ItemList) and the actual `items` (here files) in the path." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "PosixPath('/home/ubuntu/.fastai/data/mnist_tiny')" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data.path" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv')], dtype=object)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data.items" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`len(il_data)` gives you the count of files inside `il_data` and you can access individual items using index. " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "3" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "len(il_data)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "[`ItemList`](/data_block.html#ItemList) returns a single item with a single index, but returns an [`ItemList`](/data_block.html#ItemList) if given a list of indexes." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv')" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data[1]" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (1 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/labels.csv\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data[:1]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With `il_data.add` we can perform in_place concatenate another [`ItemList`](/data_block.html#ItemList) object." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (6 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/labels.csv,/home/ubuntu/.fastai/data/mnist_tiny/history.csv,/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv,/home/ubuntu/.fastai/data/mnist_tiny/labels.csv,/home/ubuntu/.fastai/data/mnist_tiny/history.csv\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data.add(il_data); il_data" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (20 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/test/1503.png,/home/ubuntu/.fastai/data/mnist_tiny/test/5071.png,/home/ubuntu/.fastai/data/mnist_tiny/test/617.png,/home/ubuntu/.fastai/data/mnist_tiny/test/585.png,/home/ubuntu/.fastai/data/mnist_tiny/test/2032.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny/test" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "itemlist = ItemList.from_folder(path_data/'test')\n", "itemlist" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As we can see, the files do not necesarily return in alpha-numeric order by default. In the above: 1503.png, ... 617.png, 585.png ...\n", "\n", "This is OK when you're always using the same machine, as the same dataset should return in the same order. But when building a datablock on one machine (say GCP) and then porting the same code to a different machine (say your laptop) that same dataset and code might return the files in a different order.\n", "\n", "Since all random operations use the loaded order of the dataset as the starting point, you will not be able to replicate any random operations, say randomly splitting the data into 80% train, and 20% validation, even while correctly seeding.\n", "\n", "The solution is to use `presort=True` in the `.from_folder()` method. As can be seen below, with that argument turned on, the file return in ascending order, and this behavior will match across machines and across platforms. Now you can reproduce any random operation you perfrom on the loaded data." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (20 items)\n", "/home/user/.fastai/data/mnist_tiny/test/1503.png,/home/user/.fastai/data/mnist_tiny/test/1605.png,/home/user/.fastai/data/mnist_tiny/test/1883.png,/home/user/.fastai/data/mnist_tiny/test/2032.png,/home/user/.fastai/data/mnist_tiny/test/205.png\n", "Path: /home/user/.fastai/data/mnist_tiny/test" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "itemlist = ItemList.from_folder(path_data/'test', presort=True)\n", "itemlist" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "How does such output above is generated?\n", "\n", "behind the scenes, executing `itemlist` calls [`ItemList.__repr__`](/data_block.html#ItemList.__repr__) which basically prints out `itemlist[0]` to `itemlist[4]`" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test/1503.png')" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "itemlist[0]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "and `itemlist[0]` basically calls `itemlist.get(0)` which returns `itemlist.items[0]`. That's why we have outputs like above." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Once you have selected the class that is suitable, you can instantiate it with one of the following factory methods" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

from_folder[source][test]

\n", "\n", "> from_folder(**`path`**:`PathOrStr`, **`extensions`**:`StrList`=***`None`***, **`recurse`**:`bool`=***`True`***, **`include`**:`OptStrList`=***`None`***, **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **`presort`**:`Optional`\\[`bool`\\]=***`False`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

Tests found for from_folder:

Some other tests where from_folder is used:

  • pytest -sv tests/test_data_block.py::test_wrong_order [source]

To run tests please refer to this guide.

\n", "\n", "Create an [`ItemList`](/data_block.html#ItemList) in `path` from the filenames that have a suffix in `extensions`. [`recurse`](/core.html#recurse) determines if we search subfolders. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.from_folder)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_TINY)\n", "path.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (1428 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_folder(path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`path` is your root data folder. In the `path` directory you have _train_ and _valid_ folders which would contain your images. For the below example, _train_ folder contains two folders/classes _cat_ and _dog_.\n", "\n", "\"from_folder\"" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

from_df[source][test]

\n", "\n", "> from_df(**`df`**:`DataFrame`, **`path`**:`PathOrStr`=***`'.'`***, **`cols`**:`IntsOrStrs`=***`0`***, **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

Tests found for from_df:

Some other tests where from_df is used:

  • pytest -sv tests/test_data_block.py::test_category [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_multi_category [source]
  • pytest -sv tests/test_data_block.py::test_regression [source]

To run tests please refer to this guide.

\n", "\n", "Create an [`ItemList`](/data_block.html#ItemList) in `path` from the inputs in the `cols` of `df`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.from_df)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Dataframe has 2 columns. The first column is the path to the image and the second column contains label id for that image. In case you have multi-labels (i.e more than one label for a single image), you will have a space(as determined by `label_delim` argument of `label_from_df`) seperated string in the labels column.\n", "\n", "`from_df` and `from_csv` can be used in a more general way. In cases you are not able to figure out how to get your ImageList, it is very easy to make a csv file with the above format.\n", "\n", "How to set `path`? `path` refers to your root data directory. So the paths in your csv file should be relative to `path` and not absolute paths. In the below example, in _labels.csv_ the paths to the images are __path + train/3/7463.png__" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_sample/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/item_list.txt'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/trained_model.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "path.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namelabel
0train/3/7463.png0
1train/3/21102.png0
2train/3/31559.png0
3train/3/46882.png0
4train/3/26209.png0
\n", "
" ], "text/plain": [ " name label\n", "0 train/3/7463.png 0\n", "1 train/3/21102.png 0\n", "2 train/3/31559.png 0\n", "3 train/3/46882.png 0\n", "4 train/3/26209.png 0" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(path/'labels.csv')\n", "df.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (14434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_df(df, path)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

from_csv[source][test]

\n", "\n", "> from_csv(**`path`**:`PathOrStr`, **`csv_name`**:`str`, **`cols`**:`IntsOrStrs`=***`0`***, **`delimiter`**:`str`=***`None`***, **`header`**:`str`=***`'infer'`***, **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

No tests found for from_csv. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create an [`ItemList`](/data_block.html#ItemList) in `path` from the inputs in the `cols` of `path/csv_name` " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.from_csv)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_sample/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/item_list.txt'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/trained_model.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_sample/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "path.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (14434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_csv(path, 'labels.csv')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Optional step: filter your data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The factory method may have grabbed too many items. For instance, if you were searching sub folders with the `from_folder` method, you may have gotten files you don't want. To remove those, you can use one of the following methods." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

filter_by_func[source][test]

\n", "\n", "> filter_by_func(**`func`**:`Callable`) → `ItemList`\n", "\n", "
×

No tests found for filter_by_func. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Only keep elements for which `func` returns `True`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.filter_by_func)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namelabel
0train/3/7463.png0
1train/3/21102.png0
2train/3/31559.png0
3train/3/46882.png0
4train/3/26209.png0
\n", "
" ], "text/plain": [ " name label\n", "0 train/3/7463.png 0\n", "1 train/3/21102.png 0\n", "2 train/3/31559.png 0\n", "3 train/3/46882.png 0\n", "4 train/3/26209.png 0" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "df = pd.read_csv(path/'labels.csv')\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Suppose that you only want to keep images with a suffix \".png\". Well, this method will do magic for you." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'.png'" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "Path(df.name[0]).suffix" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (14434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_df(df, path).filter_by_func(lambda fname: Path(fname).suffix == '.png')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

filter_by_folder[source][test]

\n", "\n", "> filter_by_folder(**`include`**=***`None`***, **`exclude`**=***`None`***)\n", "\n", "
×

Tests found for filter_by_folder:

  • pytest -sv tests/test_data_block.py::test_filter_by_folder [source]

To run tests please refer to this guide.

\n", "\n", "Only keep filenames in `include` folder or reject the ones in `exclude`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.filter_by_folder)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

filter_by_rand[source][test]

\n", "\n", "> filter_by_rand(**`p`**:`float`, **`seed`**:`int`=***`None`***)\n", "\n", "
×

No tests found for filter_by_rand. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Keep random sample of `items` with probability `p` and an optional `seed`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.filter_by_rand)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (7267 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "ImageList.from_folder(path).filter_by_rand(0.5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Contrast the number of items with the list created without the filter." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (14434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_folder(path)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

to_text[source][test]

\n", "\n", "> to_text(**`fn`**:`str`)\n", "\n", "
×

No tests found for to_text. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Save `self.items` to `fn` in `self.path`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.to_text)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namelabel
0train/3/7463.png0
1train/3/21102.png0
2train/3/31559.png0
3train/3/46882.png0
4train/3/26209.png0
\n", "
" ], "text/plain": [ " name label\n", "0 train/3/7463.png 0\n", "1 train/3/21102.png 0\n", "2 train/3/31559.png 0\n", "3 train/3/46882.png 0\n", "4 train/3/26209.png 0" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "pd.read_csv(path/'labels.csv').head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "file_name = \"item_list.txt\"\n", "ImageList.from_folder(path).to_text(file_name)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "train/3/5736.png\r\n", "train/3/35272.png\r\n", "train/3/26596.png\r\n", "train/3/42120.png\r\n", "train/3/39675.png\r\n", "train/3/47881.png\r\n", "train/3/38241.png\r\n", "train/3/59054.png\r\n", "train/3/9932.png\r\n", "train/3/50184.png\r\n", "cat: write error: Broken pipe\r\n" ] } ], "source": [ "! cat {path/file_name} | head" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

use_partial_data[source][test]

\n", "\n", "> use_partial_data(**`sample_pct`**:`float`=***`0.01`***, **`seed`**:`int`=***`None`***) → `ItemList`\n", "\n", "
×

No tests found for use_partial_data. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Use only a sample of `sample_pct`of the full dataset and an optional `seed`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.use_partial_data)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (7217 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "ImageList.from_folder(path).use_partial_data(0.5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Contrast the number of items with the list created without the filter." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (14434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ImageList.from_folder(path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Writing your own [`ItemList`](/data_block.html#ItemList)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "First check if you can't easily customize one of the existing subclass by:\n", "- subclassing an existing one and replacing the `get` method (or the `open` method if you're dealing with images)\n", "- applying a custom `processor` (see step 4)\n", "- changing the default `label_cls` for the label creation\n", "- adding a default [`PreProcessor`](/data_block.html#PreProcessor) with the `_processor` class variable\n", "\n", "If this isn't the case and you really need to write your own class, there is a [full tutorial](/tutorial.itemlist) that explains how to proceed." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

analyze_pred[source][test]

\n", "\n", "> analyze_pred(**`pred`**:`Tensor`)\n", "\n", "
×

No tests found for analyze_pred. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Called on `pred` before `reconstruct` for additional preprocessing. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.analyze_pred)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

get[source][test]

\n", "\n", "> get(**`i`**) → `Any`\n", "\n", "
×

No tests found for get. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Subclass if you want to customize how to create item `i` from `self.items`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.get)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We will have a glimpse of how `get` work with the following demo. " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (20 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/test/1503.png,/home/ubuntu/.fastai/data/mnist_tiny/test/5071.png,/home/ubuntu/.fastai/data/mnist_tiny/test/617.png,/home/ubuntu/.fastai/data/mnist_tiny/test/585.png,/home/ubuntu/.fastai/data/mnist_tiny/test/2032.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data_base = ItemList.from_folder(path=path_data, extensions=['.png'], include=['test'])\n", "il_data_base" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`get` is used inexplicitly within `il_data_base[15]`. `il_data_base.get(15)` gives the same result here, because its defulat it's to return that." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test/6736.png')" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data_base[15]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "While creating your custom [`ItemList`](/data_block.html#ItemList) however, you can override this function to do some things to your item (like opening an image)." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (20 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data_image = ImageList.from_folder(path=path_data, extensions=['.png'], include=['test'])\n", "il_data_image" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Again, normally `get` is used inexplicitly within `il_data_image[15]`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "image/jpeg": "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\n", "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAYAAAByDd+UAAAABHNCSVQICAgIfAhkiAAAAOlJREFUSIntlksOwyAMRE3Vgw0ny3Ay52Z0UVGlfE3Ssqg6EpuE5MEzjuJEJMrC3FbC/sCv5G6dCEAAvF0jKQBk3/cpaLQMVY2WkIwAeu/6LDClBTXX0Hsvzjnx3ksIYagx1z+ttDUAVHeoqteU1kAtzR3YHLAHMcLsQJJXIPPAVq1Gp/KSUpLdnRqh509oDW6AngemcQzJ7tyi8QEIyfxyM5e/pbOHIG+TKaWpJiMtAKr1Gz1XAI+rnYVZrbhElafPl+cQQuF+27ZmXZxzzXt53lY/G6PGdlv0GjtFVc+ASqUr8vt/bcuBD5ipIJ8bKsRaAAAAAElFTkSuQmCC\n", "text/plain": [ "Image (3, 28, 28)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_data_image[15]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The reason why an image is printed out instead of a FilePath object, is [`ImageList.get`](/vision.data.html#ImageList.get) overwrites [`ItemList.get`](/data_block.html#ItemList.get) and use [`ImageList.open`](/vision.data.html#ImageList.open) to print an image." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

new[source][test]

\n", "\n", "> new(**`items`**:`Iterator`\\[`T_co`\\], **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

No tests found for new. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a new [`ItemList`](/data_block.html#ItemList) from `items`, keeping the same attributes. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.new)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You'll never need to subclass this normally, just don't forget to add to `self.copy_new` the names of the arguments that needs to be copied each time `new` is called in `__init__`." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We will get a feel of how `new` works with the following examples." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (699 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7692.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7484.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9157.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/8703.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9182.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny/valid" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "itemlist1 = ItemList.from_folder(path=path_data/'valid', extensions=['.png'])\n", "itemlist1" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As you will see below, `copy_new` allows use to borrow any argument and its value from `itemlist1`, and `itemlist1.new(itemlist1.items)` allows us to use `items` and arguments inside `copy_new` to create another [`ItemList`](/data_block.html#ItemList) by calling [`ItemList.__init__`](/data_block.html#ItemList.__init__)." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "True" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "itemlist1.copy_new == ['x', 'label_cls', 'path']" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "False" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "((itemlist1.x == itemlist1.label_cls == itemlist1.inner_df == None) \n", " and (itemlist1.path == Path('/Users/Natsume/.fastai/data/mnist_tiny/valid')))\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can select any argument from [`ItemList.__init__`](/data_block.html#ItemList.__init__)'s signature and change their values. " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "itemlist1.copy_new = ['x', 'label_cls', 'path', 'inner_df']\n", "itemlist1.x = itemlist1.label_cls = itemlist1.path = itemlist1.inner_df = 'test'\n", "itemlist2 = itemlist1.new(items=itemlist1.items)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "True" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "(itemlist2.inner_df == itemlist2.x == itemlist2.label_cls == 'test' \n", "and itemlist2.path == Path('test'))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

reconstruct[source][test]

\n", "\n", "> reconstruct(**`t`**:`Tensor`, **`x`**:`Tensor`=***`None`***)\n", "\n", "
×

No tests found for reconstruct. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Reconstruct one of the underlying item for its data `t`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.reconstruct)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 2: Split the data between the training and the validation set" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This step is normally straightforward, you just have to pick one of the following functions depending on what you need." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_none[source][test]

\n", "\n", "> split_none()\n", "\n", "
×

No tests found for split_none. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Don't split the data and create an empty validation set. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_none)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_rand_pct[source][test]

\n", "\n", "> split_by_rand_pct(**`valid_pct`**:`float`=***`0.2`***, **`seed`**:`int`=***`None`***) → `ItemLists`\n", "\n", "
×

Tests found for split_by_rand_pct:

  • pytest -sv tests/test_data_block.py::test_splitdata_datasets [source]

Some other tests where split_by_rand_pct is used:

  • pytest -sv tests/test_data_block.py::test_regression [source]

To run tests please refer to this guide.

\n", "\n", "Split the items randomly by putting `valid_pct` in the validation set, optional `seed` can be passed. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_rand_pct)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_subsets[source][test]

\n", "\n", "> split_subsets(**`train_size`**:`float`, **`valid_size`**:`float`, **`seed`**=***`None`***) → `ItemLists`\n", "\n", "
×

Tests found for split_subsets:

  • pytest -sv tests/test_data_block.py::test_split_subsets [source]

To run tests please refer to this guide.

\n", "\n", "Split the items into train set with size `train_size * n` and valid set with size `valid_size * n`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_subsets)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This function is handy if you want to work with subsets of specific sizes, e.g., you want to use 20% of the data for the validation dataset, but you only want to train on a small subset of the rest of the data: `split_subsets(train_size=0.08, valid_size=0.2)`." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_files[source][test]

\n", "\n", "> split_by_files(**`valid_names`**:`ItemList`) → `ItemLists`\n", "\n", "
×

No tests found for split_by_files. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data by using the names in `valid_names` for validation. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_files)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_fname_file[source][test]

\n", "\n", "> split_by_fname_file(**`fname`**:`PathOrStr`, **`path`**:`PathOrStr`=***`None`***) → `ItemLists`\n", "\n", "
×

No tests found for split_by_fname_file. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data by using the names in `fname` for the validation set. `path` will override `self.path`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_fname_file)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Internally makes a call to `split_by_files`. `fname` contains your image file names like 0001.png." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_folder[source][test]

\n", "\n", "> split_by_folder(**`train`**:`str`=***`'train'`***, **`valid`**:`str`=***`'valid'`***) → `ItemLists`\n", "\n", "
×

Tests found for split_by_folder:

Some other tests where split_by_folder is used:

  • pytest -sv tests/test_data_block.py::test_wrong_order [source]

To run tests please refer to this guide.

\n", "\n", "Split the data depending on the folder (`train` or `valid`) in which the filenames are. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_folder)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Note: This method looks at the folder immediately after `self.path` for `valid` and `train`.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_note(\"This method looks at the folder immediately after `self.path` for `valid` and `train`.\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Basically, `split_by_folder` takes in two folder names ('train' and 'valid' in the following example), to split `il` the large [`ImageList`](/vision.data.html#ImageList) into two smaller [`ImageList`](/vision.data.html#ImageList)s, one for training set and the other for validation set. Both [`ImageList`](/vision.data.html#ImageList)s are attached to a large [`ItemLists`](/data_block.html#ItemLists) which is the final output of `split_by_folder`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/export.pkl'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/train'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv'),\n", " PosixPath('/home/ubuntu/.fastai/data/mnist_tiny/valid')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemList (1439 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/labels.csv,/home/ubuntu/.fastai/data/mnist_tiny/export.pkl,/home/ubuntu/.fastai/data/mnist_tiny/history.csv,/home/ubuntu/.fastai/data/mnist_tiny/cleaned.csv,/home/ubuntu/.fastai/data/mnist_tiny/test/1503.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il = ItemList.from_folder(path=path_data); il" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ItemList (713 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/train/export.pkl,/home/ubuntu/.fastai/data/mnist_tiny/train/3/9932.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/7189.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8498.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8888.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Valid: ItemList (699 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7692.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7484.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9157.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/8703.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9182.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd = il.split_by_folder(train='train', valid='valid'); sd" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, `split_by_folder` uses `_get_by_folder(name)`, to turn both 'train' and 'valid' folders into two list of indexes, and pass them onto `split_by_idxs` to split `il` into two [`ImageList`](/vision.data.html#ImageList)s, and finally attached to a [`ItemLists`](/data_block.html#ItemLists). " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "([24, 25, 26, 27, 28], [732, 733, 734, 735, 736], 713)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "train_idx = il._get_by_folder(name='train')\n", "train_idx[:5], train_idx[-5:], len(train_idx)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "([740, 741, 742, 743, 744], [1434, 1435, 1436, 1437, 1438], 699)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "valid_idx = il._get_by_folder(name='valid') \n", "valid_idx[:5], valid_idx[-5:],len(valid_idx)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "By the way, `_get_by_folder(name)` works in the following way, first, index the entire `il.items`, loop every item and if an item belongs to the named folder, e.g., 'train', then put it into a list. The folder `name` is the only input, and output is the list." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_idx[source][test]

\n", "\n", "> split_by_idx(**`valid_idx`**:`Collection`\\[`int`\\]) → `ItemLists`\n", "\n", "
×

Tests found for split_by_idx:

Some other tests where split_by_idx is used:

  • pytest -sv tests/test_data_block.py::test_category [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_multi_category [source]

To run tests please refer to this guide.

\n", "\n", "Split the data according to the indexes in `valid_idx`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_idx)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namelabel
0train/3/7463.png0
1train/3/21102.png0
2train/3/31559.png0
3train/3/46882.png0
4train/3/26209.png0
\n", "
" ], "text/plain": [ " name label\n", "0 train/3/7463.png 0\n", "1 train/3/21102.png 0\n", "2 train/3/31559.png 0\n", "3 train/3/46882.png 0\n", "4 train/3/26209.png 0" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "df = pd.read_csv(path/'labels.csv')\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can pass a list of indices that you want to put in the validation set like [1, 3, 10]. Or you can pass a contiguous list like `list(range(1000))`" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (13434 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample;\n", "\n", "Valid: ImageList (1000 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data = (ImageList.from_df(df, path)\n", " .split_by_idx(list(range(1000))))\n", "data" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_idxs[source][test]

\n", "\n", "> split_by_idxs(**`train_idx`**, **`valid_idx`**)\n", "\n", "
×

No tests found for split_by_idxs. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data between `train_idx` and `valid_idx`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_idxs)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, `split_by_idxs` turns two index lists (`train_idx` and `valid_idx`) into two [`ImageList`](/vision.data.html#ImageList)s, and then pass onto `split_by_list` to split `il` into two [`ImageList`](/vision.data.html#ImageList)s and attach to a [`ItemLists`](/data_block.html#ItemLists)." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ItemList (713 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/train/export.pkl,/home/ubuntu/.fastai/data/mnist_tiny/train/3/9932.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/7189.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8498.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8888.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Valid: ItemList (699 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7692.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7484.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9157.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/8703.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9182.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd = il.split_by_idxs(train_idx=train_idx, valid_idx=valid_idx); sd" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_list[source][test]

\n", "\n", "> split_by_list(**`train`**, **`valid`**)\n", "\n", "
×

No tests found for split_by_list. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data between `train` and `valid`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_list)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`split_by_list` takes in two [`ImageList`](/vision.data.html#ImageList)s which in the case below are `il[train_idx]` and `il[valid_idx]`, and pass them onto `_split` ([`ItemLists`](/data_block.html#ItemLists)) to initialize an [`ItemLists`](/data_block.html#ItemLists) object, which basically takes in the training, valiation and testing (optionally) [`ImageList`](/vision.data.html#ImageList)s as its properties." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ItemList (713 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/train/export.pkl,/home/ubuntu/.fastai/data/mnist_tiny/train/3/9932.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/7189.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8498.png,/home/ubuntu/.fastai/data/mnist_tiny/train/3/8888.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Valid: ItemList (699 items)\n", "/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7692.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/7484.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9157.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/8703.png,/home/ubuntu/.fastai/data/mnist_tiny/valid/3/9182.png\n", "Path: /home/ubuntu/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd = il.split_by_list(train=il[train_idx], valid=il[valid_idx]); sd" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This is more of an internal method, you should be using `split_by_files` if you want to pass a list of filenames for the validation set." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_by_valid_func[source][test]

\n", "\n", "> split_by_valid_func(**`func`**:`Callable`) → `ItemLists`\n", "\n", "
×

No tests found for split_by_valid_func. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data by result of `func` (which returns `True` for validation set). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_by_valid_func)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

split_from_df[source][test]

\n", "\n", "> split_from_df(**`col`**:`IntsOrStrs`=***`2`***)\n", "\n", "
×

No tests found for split_from_df. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Split the data from the `col` in the dataframe in `self.inner_df`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.split_from_df)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To use this function, you need a boolean column `is_valid`. If `is_valid[index] = True`, then that example is put in the validation set and if `is_valid[index] = False` the example is put in the training set." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(14434, 3)\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namelabelis_valid
2071train/3/28571.png0True
9382train/7/24434.png1False
6399train/7/56604.png1True
130train/3/4740.png0True
9226train/7/18876.png1False
\n", "
" ], "text/plain": [ " name label is_valid\n", "2071 train/3/28571.png 0 True\n", "9382 train/7/24434.png 1 False\n", "6399 train/7/56604.png 1 True\n", "130 train/3/4740.png 0 True\n", "9226 train/7/18876.png 1 False" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path = untar_data(URLs.MNIST_SAMPLE)\n", "df = pd.read_csv(path/'labels.csv')\n", "\n", "# Create a new column for is_valid\n", "df['is_valid'] = [True]*(df.shape[0]//2) + [False]*(df.shape[0]//2)\n", "\n", "# Randomly shuffle dataframe\n", "df = df.reindex(np.random.permutation(df.index))\n", "print(df.shape)\n", "df.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (7217 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample;\n", "\n", "Valid: ImageList (7217 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /home/ubuntu/.fastai/data/mnist_sample;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data = (ImageList.from_df(df, path)\n", " .split_from_df())\n", "data" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Warning: This method assumes the data has been created from a csv file or a dataframe.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_warn(\"This method assumes the data has been created from a csv file or a dataframe.\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 3: Label the inputs" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To label your inputs, use one of the following functions. Note that even if it's not in the documented arguments, you can always pass a `label_cls` that will be used to create those labels (the default is the one from your input [`ItemList`](/data_block.html#ItemList), and if there is none, it will go to [`CategoryList`](/data_block.html#CategoryList), [`MultiCategoryList`](/data_block.html#MultiCategoryList) or [`FloatList`](/data_block.html#FloatList) depending on the type of the labels). This is implemented in the following function:" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

get_label_cls[source][test]

\n", "\n", "> get_label_cls(**`labels`**, **`label_cls`**:`Callable`=***`None`***, **`label_delim`**:`str`=***`None`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for get_label_cls. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Return `label_cls` or guess one from the first element of `labels`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.get_label_cls)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, [`ItemList.get_label_cls`](/data_block.html#ItemList.get_label_cls) basically select a label class according to the item type of `labels`, whereas `labels` can be any of `Collection`, `pandas.core.frame.DataFrame`, `pandas.core.series.Series`. If the list elements are of type string or integer, `get_label_cls` will output [`CategoryList`](/data_block.html#CategoryList); they are of type float, then it will output [`FloatList`](/data_block.html#FloatList); if they are of type Collection, then it will output `MultiCateogryList`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "sd = ImageList.from_folder(path_data).split_by_folder('train', 'valid'); sd" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "fastai.data_block.CategoryList" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "labels = ['7', '3']\n", "label_cls = sd.train.get_label_cls(labels); label_cls" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "fastai.data_block.CategoryList" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "labels = [7, 3]\n", "label_cls = sd.train.get_label_cls(labels); label_cls" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "fastai.data_block.FloatList" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "labels = [7.0, 3.0]\n", "label_cls = sd.train.get_label_cls(labels); label_cls" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "fastai.data_block.MultiCategoryList" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "labels = [[7, 3],]\n", "label_cls = sd.train.get_label_cls(labels); label_cls" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "fastai.data_block.MultiCategoryList" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "labels = [['7', '3'],]\n", "label_cls = sd.train.get_label_cls(labels); label_cls" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "If no `label_cls` argument is passed, the correct labeling type can usually be inferred based on the data (for classification or regression). If you have multiple regression targets (e.g. predict 5 different numbers from a single image/text), be aware that arrays of floats are by default considered to be targets for one-hot encoded classification. If your task is regression, be sure the pass `label_cls = FloatList` so that learners created from your databunch initialize correctly." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first example in these docs created labels as follows:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "path = untar_data(URLs.MNIST_TINY)\n", "ll = ImageList.from_folder(path).split_by_folder().label_from_folder().train" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "If you want to save the data necessary to recreate your [`LabelList`](/data_block.html#LabelList) (not including saving the actual image/text/etc files), you can use `to_df` or `to_csv`:\n", "\n", "```python\n", "ll.train.to_csv('tmp.csv')\n", "```\n", "\n", "Or just grab a `pd.DataFrame` directly:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
xy
0train/7/9243.png7
1train/7/9519.png7
2train/7/7534.png7
3train/7/9082.png7
4train/7/8377.png7
\n", "
" ], "text/plain": [ " x y\n", "0 train/7/9243.png 7\n", "1 train/7/9519.png 7\n", "2 train/7/7534.png 7\n", "3 train/7/9082.png 7\n", "4 train/7/8377.png 7" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ll.to_df().head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_empty[source][test]

\n", "\n", "> label_empty(**\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for label_empty. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Label every item with an [`EmptyLabel`](/core.html#EmptyLabel). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_empty)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_from_df[source][test]

\n", "\n", "> label_from_df(**`cols`**:`IntsOrStrs`=***`1`***, **`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**)\n", "\n", "
×

Tests found for label_from_df:

Some other tests where label_from_df is used:

  • pytest -sv tests/test_data_block.py::test_category [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_multi_category [source]
  • pytest -sv tests/test_data_block.py::test_regression [source]

To run tests please refer to this guide.

\n", "\n", "Label `self.items` from the values in `cols` in `self.inner_df`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_from_df)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Warning: This method only works with data objects created with either `from_csv` or `from_df` methods.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_warn(\"This method only works with data objects created with either `from_csv` or `from_df` methods.\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_const[source][test]

\n", "\n", "> label_const(**`const`**:`Any`=***`0`***, **`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

Tests found for label_const:

Some other tests where label_const is used:

  • pytest -sv tests/test_data_block.py::test_split_subsets [source]
  • pytest -sv tests/test_data_block.py::test_splitdata_datasets [source]

To run tests please refer to this guide.

\n", "\n", "Label every item with `const`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_const)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_from_folder[source][test]

\n", "\n", "> label_from_folder(**`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

Tests found for label_from_folder:

  • pytest -sv tests/test_text_data.py::test_filter_classes [source]
  • pytest -sv tests/test_text_data.py::test_from_folder [source]

Some other tests where label_from_folder is used:

  • pytest -sv tests/test_data_block.py::test_wrong_order [source]

To run tests please refer to this guide.

\n", "\n", "Give a label to each filename depending on its folder. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_from_folder)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Note: This method looks at the last subfolder in the path to determine the classes.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_note(\"This method looks at the last subfolder in the path to determine the classes.\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, when an [`ItemList`](/data_block.html#ItemList) calls `label_from_folder`, it creates a lambda function which outputs a foldername which a file Path object immediately or directly belongs to, and then calls `label_from_func` with the lambda function as input. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "On the practical and high level, `label_from_folder` is mostly used with [`ItemLists`](/data_block.html#ItemLists) rather than [`ItemList`](/data_block.html#ItemList) for simplicity and efficiency, for details see the `label_from_folder` example on [ItemLists](). Even when you just want a training set [`ItemList`](/data_block.html#ItemList), you still need to do `split_none` to create an [`ItemLists`](/data_block.html#ItemLists) and then do labeling with `label_from_folder`, as the example shown below." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/valid'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: LabelList (0 items)\n", "x: ImageList\n", "\n", "y: CategoryList\n", "\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd_train = ImageList.from_folder(path_data/'train').split_none()\n", "ll_train = sd_train.label_from_folder(); ll_train" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_from_func[source][test]

\n", "\n", "> label_from_func(**`func`**:`Callable`, **`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

No tests found for label_from_func. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Apply `func` to every input to get its label. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_from_func)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Inside `label_from_func`, it applies the input `func` to every item of an [`ItemList`](/data_block.html#ItemList) and puts all the function outputs into a list, and then passes the list onto [`ItemList._label_from_list`](/data_block.html#ItemList._label_from_list). Below is a simple example of using `label_from_func`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "sd = ImageList.from_folder(path_data).split_by_folder('train', 'valid');sd" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "func=lambda o: (o.parts if isinstance(o, Path) else o.split(os.path.sep))[-2]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The lambda function above is to access the immediate foldername for a file Path object." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: LabelList (699 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ll = sd.label_from_func(func); ll" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_from_re[source][test]

\n", "\n", "> label_from_re(**`pat`**:`str`, **`full_path`**:`bool`=***`False`***, **`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

No tests found for label_from_re. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Apply the re in `pat` to determine the label of every filename. If `full_path`, search in the full name. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.label_from_re)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class CategoryList[source][test]

\n", "\n", "> CategoryList(**`items`**:`Iterator`\\[`T_co`\\], **`classes`**:`Collection`\\[`T_co`\\]=***`None`***, **`label_delim`**:`str`=***`None`***, **\\*\\*`kwargs`**) :: [`CategoryListBase`](/data_block.html#CategoryListBase)\n", "\n", "
×

No tests found for CategoryList. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Basic [`ItemList`](/data_block.html#ItemList) for single classification labels. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryList, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "[`ItemList`](/data_block.html#ItemList) suitable for storing labels in `items` belonging to `classes`. If `None` are passed, `classes` will be determined by the unique different labels. `processor` will default to [`CategoryProcessor`](/data_block.html#CategoryProcessor)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "[`CategoryList`](/data_block.html#CategoryList) uses `labels` to create an [`ItemList`](/data_block.html#ItemList) for dealing with categorical labels. Behind the scenes, [`CategoryList`](/data_block.html#CategoryList) is a subclass of [`CategoryListBase`](/data_block.html#CategoryListBase) which is a subclass of [`ItemList`](/data_block.html#ItemList). [`CategoryList`](/data_block.html#CategoryList) inherits from [`CategoryListBase`](/data_block.html#CategoryListBase) the properties such as `classes` (default as `None`), `filter_missing_y` (default as `True`), and has its own unique property `loss_func` (default as `CrossEntropyFlat()`), and its own class attribute `_processor` (default as [`CategoryProcessor`](/data_block.html#CategoryProcessor)). " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(array([1, 1, 1, 1, ..., 0, 0, 0, 0]), ['3', '7'], Category 7)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "ll = ImageList.from_folder(path_data).split_by_folder('train', 'valid').label_from_folder()\n", "ll.train.y.items, ll.train.y.classes, ll.train.y[0]" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "CategoryList (709 items)\n", "7,7,7,7,7\n", "Path: ." ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "cl = CategoryList(ll.train.y.items, ll.train.y.classes); cl" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For the behavior of printing out [`CategoryList`](/data_block.html#CategoryList) object or access an element using index, please see [`CategoryList.get`](/data_block.html#CategoryList.get) below." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, [`CategoryList.get`](/data_block.html#CategoryList.get) is used inexplicitly when printing out the [`CategoryList`](/data_block.html#CategoryList) object or `cl[idx]`. According to the source of [`CategoryList.get`](/data_block.html#CategoryList.get), each `item` is used to get its own `class`. When 'classes' is a list of strings, then elements of `items` are used as index of a list, therefore they must be integers in the range from 0 to `len(classes)-1`; if `classes` is a dictionary, then elements of `items` are used as keys, therefore they can be strings too. See examples below for details." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "CategoryList (5 items)\n", "3,7,9,7,3\n", "Path: ." ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "items = np.array([0, 1, 2, 1, 0])\n", "cl = CategoryList(items, classes=['3', '7', '9']); cl" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "CategoryList (5 items)\n", "3,7,9,7,3\n", "Path: ." ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "items = np.array(['3', '7', '9', '7', '3'])\n", "classes = {'3':3, '7':7, '9':9}\n", "cl = CategoryList(items, classes); cl" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class MultiCategoryList[source][test]

\n", "\n", "> MultiCategoryList(**`items`**:`Iterator`\\[`T_co`\\], **`classes`**:`Collection`\\[`T_co`\\]=***`None`***, **`label_delim`**:`str`=***`None`***, **`one_hot`**:`bool`=***`False`***, **\\*\\*`kwargs`**) :: [`CategoryListBase`](/data_block.html#CategoryListBase)\n", "\n", "
×

No tests found for MultiCategoryList. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Basic [`ItemList`](/data_block.html#ItemList) for multi-classification labels. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryList, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It will store list of labels in `items` belonging to `classes`. If `None` are passed, `classes` will be determined by the unique different labels. `sep` is used to split the content of `items` in a list of tags.\n", "\n", "If `one_hot=True`, the items contain the labels one-hot encoded. In this case, it is mandatory to pass a list of `classes` (as we can't use the different labels)." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class FloatList[source][test]

\n", "\n", "> FloatList(**`items`**:`Iterator`\\[`T_co`\\], **`log`**:`bool`=***`False`***, **`classes`**:`Collection`\\[`T_co`\\]=***`None`***, **\\*\\*`kwargs`**) :: [`ItemList`](/data_block.html#ItemList)\n", "\n", "
×

No tests found for FloatList. To contribute a test please refer to this guide and this discussion.

\n", "\n", "[`ItemList`](/data_block.html#ItemList) suitable for storing the floats in items for regression. Will add a `log` if this flag is `True`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(FloatList, title_level=3)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class EmptyLabelList[source][test]

\n", "\n", "> EmptyLabelList(**`items`**:`Iterator`\\[`T_co`\\], **`path`**:`PathOrStr`=***`'.'`***, **`label_cls`**:`Callable`=***`None`***, **`inner_df`**:`Any`=***`None`***, **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **`x`**:`ItemList`=***`None`***, **`ignore_empty`**:`bool`=***`False`***) :: [`ItemList`](/data_block.html#ItemList)\n", "\n", "
×

No tests found for EmptyLabelList. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Basic [`ItemList`](/data_block.html#ItemList) for dummy labels. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(EmptyLabelList, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Invisible step: preprocessing" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This isn't seen here in the API, but if you passed a `processor` (or a list of them) in your initial [`ItemList`](/data_block.html#ItemList) during step 1, it will be applied here. If you didn't pass any processor, a list of them might still be created depending on what is in the `_processor` variable of your class of items (this can be a list of [`PreProcessor`](/data_block.html#PreProcessor) classes).\n", "\n", "A processor is a transformation that is applied to all the inputs once at initialization, with a state computed on the training set that is then applied without modification on the validation set (and maybe the test set). For instance, it can be processing texts to tokenize then numericalize them. In that case we want the validation set to be numericalized with exactly the same vocabulary as the training set.\n", "\n", "Another example is in tabular data, where we fill missing values with (for instance) the median computed on the training set. That statistic is stored in the inner state of the [`PreProcessor`](/data_block.html#PreProcessor) and applied on the validation set.\n", "\n", "This is the generic class for all processors." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class PreProcessor[source][test]

\n", "\n", "> PreProcessor(**`ds`**:`Collection`\\[`T_co`\\]=***`None`***)\n", "\n", "
×

No tests found for PreProcessor. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Basic class for a processor that will be applied to items at the end of the data block API. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(PreProcessor, title_level=3)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

process_one[source][test]

\n", "\n", "> process_one(**`item`**:`Any`)\n", "\n", "
×

Tests found for process_one:

Some other tests where process_one is used:

  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]

To run tests please refer to this guide.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(PreProcessor.process_one)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Process one `item`. This method needs to be written in any subclass." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

process[source][test]

\n", "\n", "> process(**`ds`**:`Collection`\\[`T_co`\\])\n", "\n", "
×

No tests found for process. To contribute a test please refer to this guide and this discussion.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(PreProcessor.process)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`ds`: an object of [`ItemList`](/data_block.html#ItemList) \n", "Process a dataset. This default to apply `process_one` on every `item` of `ds`." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class CategoryProcessor[source][test]

\n", "\n", "> CategoryProcessor(**`ds`**:[`ItemList`](/data_block.html#ItemList)) :: [`PreProcessor`](/data_block.html#PreProcessor)\n", "\n", "
×

No tests found for CategoryProcessor. To contribute a test please refer to this guide and this discussion.

\n", "\n", "[`PreProcessor`](/data_block.html#PreProcessor) that create `classes` from `ds.items` and handle the mapping. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryProcessor, title_level=3)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

generate_classes[source][test]

\n", "\n", "> generate_classes(**`items`**)\n", "\n", "
×

No tests found for generate_classes. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Generate classes from `items` by taking the sorted unique values. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryProcessor.generate_classes)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

process[source][test]

\n", "\n", "> process(**`ds`**)\n", "\n", "
×

No tests found for process. To contribute a test please refer to this guide and this discussion.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryProcessor.process)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`ds` is an object of [`CategoryList`](/data_block.html#CategoryList). \n", "It basically generates a list of unique labels (assigned to `ds.classes`) and a dictionary mapping `classes` to indexes (assigned to `ds.c2i`)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It is an internal function only called to apply processors to training, validation and testing datasets after the labeling step." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class MultiCategoryProcessor[source][test]

\n", "\n", "> MultiCategoryProcessor(**`ds`**:[`ItemList`](/data_block.html#ItemList), **`one_hot`**:`bool`=***`False`***) :: [`CategoryProcessor`](/data_block.html#CategoryProcessor)\n", "\n", "
×

No tests found for MultiCategoryProcessor. To contribute a test please refer to this guide and this discussion.

\n", "\n", "[`PreProcessor`](/data_block.html#PreProcessor) that create `classes` from `ds.items` and handle the mapping. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryProcessor, title_level=3)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

generate_classes[source][test]

\n", "\n", "> generate_classes(**`items`**)\n", "\n", "
×

No tests found for generate_classes. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Generate classes from `items` by taking the sorted unique values. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryProcessor.generate_classes)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Optional steps" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Add transforms" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Transforms differ from processors in the sense they are applied on the fly when we grab one item. They also may change each time we ask for the same item in the case of random transforms." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

transform[source][test]

\n", "\n", "> transform(**`tfms`**:`Optional`\\[`Tuple`\\[`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\], `Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\]\\]\\]=***`(None, None)`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for transform. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Set `tfms` to be applied to the xs of the train and validation set. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.transform)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This is primary for the vision application. The `kwargs` arguments are the ones expected by the type of transforms you pass. `tfm_y` is among them and if set to `True`, the transforms will be applied to input and target.\n", "\n", "For examples see: [vision.transforms](vision.transform.html)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Add a test set" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To add a test set, you can use one of the two following methods." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

add_test[source][test]

\n", "\n", "> add_test(**`items`**:`Iterator`\\[`T_co`\\], **`label`**:`Any`=***`None`***)\n", "\n", "
×

No tests found for add_test. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Add test set containing `items` with an arbitrary `label`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.add_test)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Note: Here `items` can be an `ItemList` or a collection.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_note(\"Here `items` can be an `ItemList` or a collection.\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

add_test_folder[source][test]

\n", "\n", "> add_test_folder(**`test_folder`**:`str`=***`'test'`***, **`label`**:`Any`=***`None`***)\n", "\n", "
×

No tests found for add_test_folder. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Add test set containing items from `test_folder` and an arbitrary `label`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.add_test_folder)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "
Warning: In fastai the test set is unlabeled! No labels will be collected even if they are available.
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "jekyll_warn(\"In fastai the test set is unlabeled! No labels will be collected even if they are available.\")" ] }, { "cell_type": "markdown", "metadata": { "hide_input": true }, "source": [ "Instead, either the passed `label` argument or an empty label will be used for all entries of this dataset (this is required by the internal pipeline of fastai). \n", "\n", "In the `fastai` framework `test` datasets have no labels - this is the unknown data to be predicted. If you want to validate your model on a `test` dataset with labels, you probably need to use it as a validation set, as in:\n", "\n", "```\n", "data_test = (ImageList.from_folder(path)\n", " .split_by_folder(train='train', valid='test')\n", " .label_from_folder()\n", " ...)\n", "```\n", "\n", "Another approach, where you do use a normal validation set, and then when the training is over, you just want to validate the test set w/ labels as a validation set, you can do this:\n", "\n", "```\n", "tfms = []\n", "path = Path('data').resolve()\n", "data = (ImageList.from_folder(path)\n", " .split_by_pct()\n", " .label_from_folder()\n", " .transform(tfms)\n", " .databunch()\n", " .normalize() ) \n", "learn = cnn_learner(data, models.resnet50, metrics=accuracy)\n", "learn.fit_one_cycle(5,1e-2)\n", "\n", "# now replace the validation dataset entry with the test dataset as a new validation dataset: \n", "# everything is exactly the same, except replacing `split_by_pct` w/ `split_by_folder` \n", "# (or perhaps you were already using the latter, so simply switch to valid='test')\n", "data_test = (ImageList.from_folder(path)\n", " .split_by_folder(train='train', valid='test')\n", " .label_from_folder()\n", " .transform(tfms)\n", " .databunch()\n", " .normalize()\n", " ) \n", "learn.validate(data_test.valid_dl)\n", "```\n", "Of course, your data block can be totally different, this is just an example." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 4: convert to a [`DataBunch`](/basic_data.html#DataBunch)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This last step is usually pretty straightforward. You just have to include all the arguments we pass to [`DataBunch.create`](/basic_data.html#DataBunch.create) (`bs`, `num_workers`, `collate_fn`). The class called to create a [`DataBunch`](/basic_data.html#DataBunch) is set in the `_bunch` attribute of the inputs of the training set if you need to modify it. Normally, the various subclasses we showed before handle that for you." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

databunch[source][test]

\n", "\n", "> databunch(**`path`**:`PathOrStr`=***`None`***, **`bs`**:`int`=***`64`***, **`val_bs`**:`int`=***`None`***, **`num_workers`**:`int`=***`4`***, **`dl_tfms`**:`Optional`\\[`Collection`\\[`Callable`\\]\\]=***`None`***, **`device`**:[`device`](https://pytorch.org/docs/stable/tensor_attributes.html#torch-device)=***`None`***, **`collate_fn`**:`Callable`=***`'data_collate'`***, **`no_check`**:`bool`=***`False`***, **\\*\\*`kwargs`**) → `DataBunch`\n", "\n", "
×

Tests found for databunch:

  • pytest -sv tests/test_vision_data.py::test_vision_datasets [source]

Some other tests where databunch is used:

  • pytest -sv tests/test_data_block.py::test_regression [source]

To run tests please refer to this guide.

\n", "\n", "Create an [`DataBunch`](/basic_data.html#DataBunch) from self, `path` will override `self.path`, `kwargs` are passed to [`DataBunch.create`](/basic_data.html#DataBunch.create). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.databunch)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Inner classes" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class LabelList[source][test]

\n", "\n", "> LabelList(**`x`**:[`ItemList`](/data_block.html#ItemList), **`y`**:[`ItemList`](/data_block.html#ItemList), **`tfms`**:`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\]=***`None`***, **`tfm_y`**:`bool`=***`False`***, **\\*\\*`kwargs`**) :: [`Dataset`](https://pytorch.org/docs/stable/data.html#torch.utils.data.Dataset)\n", "\n", "
×

No tests found for LabelList. To contribute a test please refer to this guide and this discussion.

\n", "\n", "A list of inputs `x` and labels `y` with optional `tfms`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Optionally apply `tfms` to `y` if `tfm_y` is `True`. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, it takes inputs [`ItemList`](/data_block.html#ItemList) and labels [`ItemList`](/data_block.html#ItemList) as its properties `x` and `y`, sets property `item` to `None`, and uses [`LabelList.transform`](/data_block.html#LabelList.transform) to apply a list of transforms `TfmList` to `x` and `y` if `tfm_y` is set `True`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(ImageList (709 items)\n", " Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", " Path: /Users/Natsume/.fastai/data/mnist_tiny, CategoryList (709 items)\n", " 7,7,7,7,7\n", " Path: /Users/Natsume/.fastai/data/mnist_tiny)" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "ll = ImageList.from_folder(path_data).split_by_folder('train', 'valid').label_from_folder()\n", "ll.train.x, ll.train.y" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "LabelList(x=ll.train.x, y=ll.train.y)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

export[source][test]

\n", "\n", "> export(**`fn`**:`PathOrStr`, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for export. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Export the minimal state and save it in `fn` to load an empty version for inference. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.export)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

transform_y[source][test]

\n", "\n", "> transform_y(**`tfms`**:`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\]=***`None`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for transform_y. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Set `tfms` to be applied to the targets only. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.transform_y)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

get_state[source][test]

\n", "\n", "> get_state(**\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for get_state. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Return the minimal state for export. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.get_state)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

load_empty[source][test]

\n", "\n", "> load_empty(**`path`**:`PathOrStr`, **`fn`**:`PathOrStr`)\n", "\n", "
×

No tests found for load_empty. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Load the state in `fn` to create an empty [`LabelList`](/data_block.html#LabelList) for inference. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.load_empty)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

load_state[source][test]

\n", "\n", "> load_state(**`path`**:`PathOrStr`, **`state`**:`dict`) → `LabelList`\n", "\n", "
×

No tests found for load_state. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a [`LabelList`](/data_block.html#LabelList) from `state`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.load_state)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

process[source][test]

\n", "\n", "> process(**`xp`**:[`PreProcessor`](/data_block.html#PreProcessor)=***`None`***, **`yp`**:[`PreProcessor`](/data_block.html#PreProcessor)=***`None`***, **`name`**:`str`=***`None`***)\n", "\n", "
×

No tests found for process. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Launch the processing on `self.x` and `self.y` with `xp` and `yp`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.process)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, [`LabelList.process`](/data_block.html#LabelList.process) does 3 three things: 1. ask labels `y` to be processed by `yp` with `y.process(yp)`; 2. if `y.filter_missing_y` is `True`, then removes the missing data samples from `x` and `y`; 3. ask inputs `x` to be processed by `xp` with `x.process(xp)`" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "sd = ImageList.from_folder(path_data).split_by_folder('train', 'valid')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "sd.train = sd.train.label_from_folder(from_item_lists=True)\n", "sd.valid = sd.valid.label_from_folder(from_item_lists=True)\n", "sd.__class__ = LabelLists" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "([], [])" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "xp,yp = sd.get_processors()\n", "xp,yp" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd.train.process(xp, yp)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

set_item[source][test]

\n", "\n", "> set_item(**`item`**)\n", "\n", "
×

No tests found for set_item. To contribute a test please refer to this guide and this discussion.

\n", "\n", "For inference, will briefly replace the dataset with one that only contains `item`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.set_item)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

to_df[source][test]

\n", "\n", "> to_df()\n", "\n", "
×

No tests found for to_df. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create `pd.DataFrame` containing `items` from `self.x` and `self.y`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.to_df)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

to_csv[source][test]

\n", "\n", "> to_csv(**`dest`**:`str`)\n", "\n", "
×

No tests found for to_csv. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Save `self.to_df()` to a CSV file in `self.path`/`dest`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.to_csv)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

transform[source][test]

\n", "\n", "> transform(**`tfms`**:`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\], **`tfm_y`**:`bool`=***`None`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for transform. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Set the `tfms` and `tfm_y` value to be applied to the inputs and targets. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.transform)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class ItemLists[source][test]

\n", "\n", "> ItemLists(**`path`**:`PathOrStr`, **`train`**:[`ItemList`](/data_block.html#ItemList), **`valid`**:[`ItemList`](/data_block.html#ItemList))\n", "\n", "
×

No tests found for ItemLists. To contribute a test please refer to this guide and this discussion.

\n", "\n", "An [`ItemList`](/data_block.html#ItemList) for each of `train` and `valid` (optional `test`). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemLists, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It initializes an [`ItemLists`](/data_block.html#ItemLists) object, which basically brings in the training, valiation and testing (optionally) [`ItemList`](/data_block.html#ItemList)s as its properties. It also offers helpful warning messages on situations when the training or validation [`ItemList`](/data_block.html#ItemList) is empty. \n", "\n", "See the following example for how to create an [`ItemLists`](/data_block.html#ItemLists) object. " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/valid'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "il_train = ImageList.from_folder(path_data/'train')\n", "il_valid = ImageList.from_folder(path_data/'valid')\n", "il_test = ImageList.from_folder(path_data/'test')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/valid;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ils = ItemLists(path=path_data, train=il_train, valid=il_valid); ils" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/valid;\n", "\n", "Test: ImageList (20 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/test" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ils.test = il_test; ils" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "However, we are most likely to see an [`ItemLists`](/data_block.html#ItemLists), right after a large [`ItemList`](/data_block.html#ItemList) is splitted and turned into an [`ItemLists`](/data_block.html#ItemLists) by methods like [`ItemList.split_by_folder`](/data_block.html#ItemList.split_by_folder). Then, we will add labels to all training and validation simply using `sd.label_from_folder()` (`sd` is an [`ItemLists`](/data_block.html#ItemLists), see example below). Now, some of you may be surprised because `label_from_folder` is a method of [`ItemList`](/data_block.html#ItemList) not [`ItemLists`](/data_block.html#ItemLists). Well, this is a magic of fastai data_block api.\n", "\n", "With the following example, we may understand a little better how to get labelling done by calling [`ItemLists.__getattr__`](/data_block.html#ItemLists.__getattr__) with [`ItemList.label_from_folder`](/data_block.html#ItemList.label_from_folder)." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ImageList (1428 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il = ImageList.from_folder(path_data); il" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "An [`ItemList`](/data_block.html#ItemList) or its subclass object must do a split to turn itself into an [`ItemLists`](/data_block.html#ItemLists) before doing labeling to become a [`LabelLists`](/data_block.html#LabelLists) object." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "sd = il.split_by_folder(train='train', valid='valid'); sd\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: LabelList (699 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ll = sd.label_from_folder(); ll" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Even when there is just an [`ImageList`](/vision.data.html#ImageList) from a traning set folder with no split needed, we still must do `split_none()` in order to create an [`ItemLists`](/data_block.html#ItemLists), and only then we can do `ItemLists.label_from_folder()` nicely." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "ItemLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: ImageList (0 items)\n", "\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "il_train = ImageList.from_folder(path_data/'train')\n", "sd_train = il_train.split_none(); sd_train" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: LabelList (0 items)\n", "x: ImageList\n", "\n", "y: CategoryList\n", "\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ll_valid_empty = sd_train.label_from_folder(); ll_valid_empty" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "So practially, although `label_from_folder` is not an [`ItemLists`](/data_block.html#ItemLists) method, we can call `ItemLists.label_from_folder()` to label training, validation and test [`ItemList`](/data_block.html#ItemList)s once for all." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, `ItemLists.label_from_folder()` actually calls `ItemLists.__getattr__('label_from_folder')`, in which all training, validation even testing [`ItemList`](/data_block.html#ItemList) get to call `label_from_folder`, and then turns the [`ItemLists`](/data_block.html#ItemLists) into a [`LabelLists`](/data_block.html#LabelLists) and calls [`LabelLists.process`](/data_block.html#LabelLists.process) at last.\n", "\n", "You can directly use `LabelLists.__getattr__` to do labelling as below." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: LabelList (709 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Valid: LabelList (699 items)\n", "x: ImageList\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "y: CategoryList\n", "7,7,7,7,7\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny;\n", "\n", "Test: None" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ld_inner = sd.__getattr__('label_from_folder'); ld_inner()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

label_from_lists[source][test]

\n", "\n", "> label_from_lists(**`train_labels`**:`Iterator`\\[`T_co`\\], **`valid_labels`**:`Iterator`\\[`T_co`\\], **`label_cls`**:`Callable`=***`None`***, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

No tests found for label_from_lists. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Use the labels in `train_labels` and `valid_labels` to label the data. `label_cls` will overwrite the default. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemLists.label_from_lists)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

transform[source][test]

\n", "\n", "> transform(**`tfms`**:`Optional`\\[`Tuple`\\[`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\], `Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\]\\]\\]=***`(None, None)`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for transform. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Set `tfms` to be applied to the xs of the train and validation set. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemLists.transform)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

transform_y[source][test]

\n", "\n", "> transform_y(**`tfms`**:`Optional`\\[`Tuple`\\[`Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\], `Union`\\[`Callable`, `Collection`\\[`Callable`\\]\\]\\]\\]=***`(None, None)`***, **\\*\\*`kwargs`**)\n", "\n", "
×

No tests found for transform_y. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Set `tfms` to be applied to the ys of the train and validation set. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemLists.transform_y)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

class LabelLists[source][test]

\n", "\n", "> LabelLists(**`path`**:`PathOrStr`, **`train`**:[`ItemList`](/data_block.html#ItemList), **`valid`**:[`ItemList`](/data_block.html#ItemList)) :: [`ItemLists`](/data_block.html#ItemLists)\n", "\n", "
×

No tests found for LabelLists. To contribute a test please refer to this guide and this discussion.

\n", "\n", "A [`LabelList`](/data_block.html#LabelList) for each of `train` and `valid` (optional `test`). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists, title_level=3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Creating a [`LabelLists`](/data_block.html#LabelLists) object is exactly the same way as creating an [`ItemLists`](/data_block.html#ItemLists) object, because its base class is [`ItemLists`](/data_block.html#ItemLists) and does not overwrite [`ItemLists.__init__`](/data_block.html#ItemLists.__init__). The example below shows how to build a [`LabelLists`](/data_block.html#LabelLists) object." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/valid'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data = untar_data(URLs.MNIST_TINY); path_data.ls()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "il_train = ImageList.from_folder(path_data/'train')\n", "il_valid = ImageList.from_folder(path_data/'valid')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LabelLists;\n", "\n", "Train: ImageList (709 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/train;\n", "\n", "Valid: ImageList (699 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/valid;\n", "\n", "Test: ImageList (20 items)\n", "Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28),Image (3, 28, 28)\n", "Path: /Users/Natsume/.fastai/data/mnist_tiny/test" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ll_test = LabelLists(path_data, il_train, il_valid); \n", "ll_test.test = il_valid = ImageList.from_folder(path_data/'test')\n", "ll_test" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

get_processors[source][test]

\n", "\n", "> get_processors()\n", "\n", "
×

No tests found for get_processors. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Read the default class processors if none have been set. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.get_processors)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Behind the scenes, `LabelLists.get_processors()` first puts `train.x._processor` classes and `train.y._processor` classes into separate lists, and then instantiates those processors and put them into `xp` and `yp`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "from fastai.vision import *" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "path_data = untar_data(URLs.MNIST_TINY)\n", "sd = ImageList.from_folder(path_data).split_by_folder('train', 'valid')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "sd.train = sd.train.label_from_folder(from_item_lists=True)\n", "sd.valid = sd.valid.label_from_folder(from_item_lists=True)\n", "sd.__class__ = LabelLists" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "xp,yp = sd.get_processors()\n", "xp,yp" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

load_empty[source][test]

\n", "\n", "> load_empty(**`path`**:`PathOrStr`, **`fn`**:`PathOrStr`=***`'export.pkl'`***)\n", "\n", "
×

No tests found for load_empty. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a [`LabelLists`](/data_block.html#LabelLists) with empty sets from the serialized file in `path/fn`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.load_empty)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

load_state[source][test]

\n", "\n", "> load_state(**`path`**:`PathOrStr`, **`state`**:`dict`)\n", "\n", "
×

No tests found for load_state. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a [`LabelLists`](/data_block.html#LabelLists) with empty sets from the serialized `state`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.load_state)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

process[source][test]

\n", "\n", "> process()\n", "\n", "
×

No tests found for process. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Process the inner datasets. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelLists.process)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": false }, "outputs": [ { "data": { "text/markdown": [ "

process[source][test]

\n", "\n", "> process(**`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***)\n", "\n", "
×

No tests found for process. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Apply `processor` or `self.processor` to `self`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.process)" ] }, { "cell_type": "markdown", "metadata": { "hide_input": true }, "source": [ "`processor` is one or more `PreProcessors` objects \n", "Behind the scenes, we put all of `processor` into a list and apply them all to an object of [`ItemList`](/data_block.html#ItemList) or its subclasses." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Helper functions" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

get_files[source][test]

\n", "\n", "> get_files(**`path`**:`PathOrStr`, **`extensions`**:`StrList`=***`None`***, **`recurse`**:`bool`=***`False`***, **`include`**:`OptStrList`=***`None`***, **`presort`**:`bool`=***`False`***) → `FilePathList`\n", "\n", "
×

No tests found for get_files. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Return list of files in `path` that have a suffix in `extensions`; optionally [`recurse`](/core.html#recurse). " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(get_files)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To to more precise, this function returns list of FilePath objects using files in `path` that must have a suffix in `extensions`, and hidden folders and files are ignored. If `recurse=True`, all files in subfolders will be applied; `include` is used to select particular folders to apply.\n", "\n", "Inside [`get_files`](/data_block.html#get_files), there is [`_get_files`](/data_block.html#_get_files) which turns all filenames inside `f` from directory `parent/p` into a list of FilePath objects. All filenames must have a suffix in `extensions`. All hidden files are ignored." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "path_data = untar_data(URLs.MNIST_TINY) " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/valid'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/models'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "path_data.ls()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With `recurse=False`, no subfolder files are made available." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_noRecurse = get_files(path_data) \n", "list_FilePath_noRecurse" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With `recurse=True`, all subfolder files are made available, except hidden files." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/valid/7/9294.png')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_recurse = get_files(path_data, recurse=True)\n", "list_FilePath_recurse[:3]" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train/3/7263.png'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/train/3/7288.png')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_recurse[-2:]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With `extensions=['.csv']`, only files with the suffix of `.csv` are made available." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/labels.csv'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/history.csv')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_recurse_csv = get_files(path_data, recurse=True, extensions=['.csv'])\n", "list_FilePath_recurse_csv" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "With `include=['test']`, only files in `path_data` and its subfolder `test` are made available." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/4605.png'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/617.png'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/205.png')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_include = get_files(path_data, recurse=True, extensions=['.png','.jpg','.jpeg'],\n", " include=['test'])\n", "list_FilePath_include[:3]" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/1605.png'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/2642.png'),\n", " PosixPath('/Users/Natsume/.fastai/data/mnist_tiny/test/5071.png')]" ] }, "execution_count": null, "metadata": {}, "output_type": "execute_result" } ], "source": [ "list_FilePath_include[-3:]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Undocumented Methods - Methods moved below this line will intentionally be hidden" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

new[source][test]

\n", "\n", "> new(**`items`**:`Iterator`\\[`T_co`\\], **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

No tests found for new. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a new [`ItemList`](/data_block.html#ItemList) from `items`, keeping the same attributes. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryList.new)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

new[source][test]

\n", "\n", "> new(**`x`**, **`y`**, **\\*\\*`kwargs`**) → `LabelList`\n", "\n", "
×

No tests found for new. To contribute a test please refer to this guide and this discussion.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.new)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

get[source][test]

\n", "\n", "> get(**`i`**)\n", "\n", "
×

No tests found for get. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Subclass if you want to customize how to create item `i` from `self.items`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryList.get)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

predict[source][test]

\n", "\n", "> predict(**`res`**)\n", "\n", "
×

No tests found for predict. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Delegates predict call on `res` to `self.y`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.predict)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

new[source][test]

\n", "\n", "> new(**`items`**:`Iterator`\\[`T_co`\\], **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

No tests found for new. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a new [`ItemList`](/data_block.html#ItemList) from `items`, keeping the same attributes. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.new)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

process_one[source][test]

\n", "\n", "> process_one(**`item`**:[`ItemBase`](/core.html#ItemBase), **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***)\n", "\n", "
×

Tests found for process_one:

Some other tests where process_one is used:

  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]

To run tests please refer to this guide.

\n", "\n", "Apply `processor` or `self.processor` to `item`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.process_one)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

process_one[source][test]

\n", "\n", "> process_one(**`item`**)\n", "\n", "
×

Tests found for process_one:

Some other tests where process_one is used:

  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]

To run tests please refer to this guide.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryProcessor.process_one)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

get[source][test]

\n", "\n", "> get(**`i`**)\n", "\n", "
×

No tests found for get. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Subclass if you want to customize how to create item `i` from `self.items`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(FloatList.get)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

process_one[source][test]

\n", "\n", "> process_one(**`item`**)\n", "\n", "
×

Tests found for process_one:

  • pytest -sv tests/test_data_block.py::test_category_processor_existing_class [source]
  • pytest -sv tests/test_data_block.py::test_category_processor_non_existing_class [source]

To run tests please refer to this guide.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryProcessor.process_one)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It basically converts `item` which is a category name to an index." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`classes`: a list of unique and sorted labels; \n", "It creates the inner mapping from category name to index (stored in `c2i`) from the `classes`." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

create_classes[source][test]

\n", "\n", "> create_classes(**`classes`**)\n", "\n", "
×

No tests found for create_classes. To contribute a test please refer to this guide and this discussion.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryProcessor.create_classes)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

get[source][test]

\n", "\n", "> get(**`i`**)\n", "\n", "
×

No tests found for get. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Subclass if you want to customize how to create item `i` from `self.items`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryList.get)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

new[source][test]

\n", "\n", "> new(**`items`**:`Iterator`\\[`T_co`\\], **`processor`**:`Union`\\[[`PreProcessor`](/data_block.html#PreProcessor), `Collection`\\[[`PreProcessor`](/data_block.html#PreProcessor)\\]\\]=***`None`***, **\\*\\*`kwargs`**) → `ItemList`\n", "\n", "
×

No tests found for new. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Create a new [`ItemList`](/data_block.html#ItemList) from `items`, keeping the same attributes. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(FloatList.new)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

reconstruct[source][test]

\n", "\n", "> reconstruct(**`t`**)\n", "\n", "
×

No tests found for reconstruct. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Reconstruct one of the underlying item for its data `t`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(FloatList.reconstruct)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

analyze_pred[source][test]

\n", "\n", "> analyze_pred(**`pred`**, **`thresh`**:`float`=***`0.5`***)\n", "\n", "
×

No tests found for analyze_pred. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Called on `pred` before `reconstruct` for additional preprocessing. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryList.analyze_pred)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

reconstruct[source][test]

\n", "\n", "> reconstruct(**`t`**)\n", "\n", "
×

No tests found for reconstruct. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Reconstruct one of the underlying item for its data `t`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(MultiCategoryList.reconstruct)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

reconstruct[source][test]

\n", "\n", "> reconstruct(**`t`**)\n", "\n", "
×

No tests found for reconstruct. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Reconstruct one of the underlying item for its data `t`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryList.reconstruct)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

analyze_pred[source][test]

\n", "\n", "> analyze_pred(**`pred`**, **`thresh`**:`float`=***`0.5`***)\n", "\n", "
×

No tests found for analyze_pred. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Called on `pred` before `reconstruct` for additional preprocessing. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(CategoryList.analyze_pred)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

reconstruct[source][test]

\n", "\n", "> reconstruct(**`t`**:`Tensor`, **`x`**:`Tensor`=***`None`***)\n", "\n", "
×

No tests found for reconstruct. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Reconstruct one of the underlying item for its data `t`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(EmptyLabelList.reconstruct)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

get[source][test]

\n", "\n", "> get(**`i`**)\n", "\n", "
×

No tests found for get. To contribute a test please refer to this guide and this discussion.

\n", "\n", "Subclass if you want to customize how to create item `i` from `self.items`. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(EmptyLabelList.get)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "

databunch[source][test]

\n", "\n", "> databunch(**\\*\\*`kwargs`**)\n", "\n", "
×

Tests found for databunch:

Some other tests where databunch is used:

  • pytest -sv tests/test_data_block.py::test_regression [source]

To run tests please refer to this guide.

\n", "\n", "To throw a clear error message when the data wasn't split. " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(LabelList.databunch)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## New Methods - Please document or move to the undocumented section" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "hide_input": true }, "outputs": [ { "data": { "text/markdown": [ "

add[source][test]

\n", "\n", "> add(**`items`**:`ItemList`)\n", "\n", "
×

No tests found for add. To contribute a test please refer to this guide and this discussion.

" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "show_doc(ItemList.add)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [] } ], "metadata": { "jekyll": { "keywords": "fastai", "summary": "The data block API", "title": "data_block" }, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" } }, "nbformat": 4, "nbformat_minor": 2 }