{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "This notebook is about Data Visualisation of high dimensional data to a lower dimension using Tensorboard with t-SNE and PCA dimensionality reduction techniques and exploration of data points with multiple parameter tuning.\n", "\n", "## What is an Embedding?\n", "* Embedding is a way to map discrete objects (images, words, etc.) to high dimensional vectors.\n", "* The individual dimensions in these vectors typically have no inherent meaning. Instead, it's the overall patterns of location and distance between vectors that machine learning takes advantage of.\n", "* Embeddings are important for input to machine learning. Classifiers, and neural networks more generally, work on vectors of real numbers. They train best on dense vectors, where all values contribute to define an object.\n", "* Embeddings map objects to vectors, applications can use similarity in vector space (for instance, Euclidean distance or the angle between vectors) as a robust and flexible measure of object similarity. One common use is to find nearest neighbors.\n", "\n", "## Workflow\n", "To do a basic embedding visualisation, I'll follow the below workflow steps:\n", "1. Read the Fashion MNIST data and create an X (image) and Y (label) batch\n", "2. Create a Summary Writer\n", "3. Configure the projector\n", "4. Create the embedding Tensor from X\n", "5. Run the TF session and create a model check-point\n", "6. Create the sprite image\n", "7. Create the metadata (labels) file\n", "\n", "But first, let's import the necessary libraries:" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['.DS_Store', 'fashion-mnist_test.csv', 'fashion-mnist_train.csv', 't10k-images-idx3-ubyte', 't10k-labels-idx1-ubyte', 'train-images-idx3-ubyte', 'train-labels-idx1-ubyte']\n" ] } ], "source": [ "# Import Numpy for statistical calculations\n", "import numpy as np\n", "\n", "# Import Pandas for data manipulation using dataframes\n", "import pandas as pd\n", "\n", "# Import Warnings \n", "import warnings\n", "warnings.filterwarnings('ignore')\n", "from IPython.display import Image\n", "from IPython.core.display import HTML \n", "\n", "# Import matplotlib Library for data visualisation\n", "import matplotlib.pyplot as plt\n", "%matplotlib inline\n", "import os\n", "import tensorflow as tf\n", "from tensorflow.contrib.tensorboard.plugins import projector\n", "from tensorflow.examples.tutorials.mnist import input_data\n", "\n", "# Input data files are available in the \"data/\" directory.\n", "# For example, running this will list the files in the input directory\n", "import os\n", "print(os.listdir(\"data\"))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**1 - Read the Fashion MNIST test data and create an X (image) and Y (label) batch**\n", "\n", "I will load the `fashion-mnist` test data into a pandas dataframe and subsequently convert it into a numpy array with datatype as float32." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": true }, "outputs": [], "source": [ "test_data = np.array(pd.read_csv('data/fashion-mnist_test.csv'), dtype='float32')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's consider around 2500 images as part of the embedding." ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "collapsed": true }, "outputs": [], "source": [ "embed_count = 2500" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now I will split the data into **X_test** for storing images and **Y_test** for storing labels. Since all images are ranging from 0-255, I will need to rescale all of them by dividing with 255 so that it reflects between 0 and 1." ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "collapsed": true }, "outputs": [], "source": [ "X_test = test_data[:embed_count, 1:] / 255\n", "Y_test = test_data[:embed_count, 0]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**2 - Create a Summary Writer**\n", "\n", "To store all the events in a log directory for tensorboard view, I need to mention the path as below." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "collapsed": true }, "outputs": [], "source": [ "logdir = 'fashionMNIST-logs'\n", "\n", "# Use this logdir to create a summary writer\n", "summary_writer = tf.summary.FileWriter(logdir)" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# Creating the embedding variable with all the images defined above under X_test\n", "embedding_var = tf.Variable(X_test, name='fmnist_embedding')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**3 - Configure the projector**\n", "\n", "This is the important part of embedding visualisation. Here I specify what variable I want for the project, what the metadata path is (the names and classes), and where to save the sprites." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# Format: tensorflow/contrib/tensorboard/plugins/projector/projector_config.proto\n", "config = projector.ProjectorConfig()\n", "\n", "# You can add multiple embeddings. Here I add only one.\n", "embedding = config.embeddings.add()\n", "embedding.tensor_name = embedding_var.name\n", "\n", "# Link this tensor to its metadata file (e.g. labels).\n", "embedding.metadata_path = os.path.join(logdir, 'metadata.tsv')\n", "\n", "# After constructing the sprite, I need to tell the Embedding Projector where to find it\n", "embedding.sprite.image_path = os.path.join(logdir, 'sprite.png')\n", "embedding.sprite.single_image_dim.extend([28, 28])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**4 - Create the embedding Tensor from X**" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# The next line writes a projector_config.pbtxt in the logdir. TensorBoard will read this file during startup.\n", "projector.visualize_embeddings(summary_writer,config)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**5 - Run the TF session and create a model check-point**" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# Periodically save the model variables in a checkpoint in logdir.\n", "with tf.Session() as sesh:\n", " sesh.run(tf.global_variables_initializer())\n", " saver = tf.train.Saver()\n", " saver.save(sesh, os.path.join(logdir, 'model.ckpt'))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**6 - Create the sprite image**\n", "\n", "This is used for making sprite images for the [embedding projector](http://projector.tensorflow.org/)." ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "collapsed": true }, "outputs": [], "source": [ "rows = 28 \n", "cols = 28\n", "label = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']\n", "\n", "sprite_dim = int(np.sqrt(X_test.shape[0]))\n", "sprite_image = np.ones((cols * sprite_dim, rows * sprite_dim))\n", "\n", "index = 0 \n", "labels = [] \n", "for i in range(sprite_dim): \n", " for j in range(sprite_dim):\n", " labels.append(label[int(Y_test[index])])\n", "\n", " sprite_image[\n", " i * cols: (i + 1) * cols,\n", " j * rows: (j + 1) * rows\n", " ] = X_test[index].reshape(28, 28) * -1 + 1\n", "\n", " index += 1" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**7 - Create the metadata (labels) file**\n", "\n", "Usually embeddings have metadata associated with it (e.g. labels, images). The metadata should be stored in a separate file outside of the model checkpoint, since the metadata is not a trainable parameter of the model." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with open(embedding.metadata_path, 'w') as meta:\n", " meta.write('Index\\tLabel\\n')\n", " for index, label in enumerate(labels):\n", " meta.write('{}\\t{}\\n'.format(index, label))\n", " \n", "plt.imsave(embedding.sprite.image_path, sprite_image, cmap='gray')\n", "plt.imshow(sprite_image, cmap='gray')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The above figure represents the grayscale fashion products representation in lower dimension.Let us look more closer by zooming the picture as below:\n", "\n", "![zoom-in](images/zoom-in-pic.png)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Launching TensorBoard\n", "To run TensorBoard, use the following command:\n", "\n", "`tensorboard --logdir=path/to/log-directory`\n", "\n", "where `logdir` points to the directory where the `FileWriter` serialized its data. If this `logdir` directory contains subdirectories which contain serialized data from separate runs, then TensorBoard will visualize the data from all of those runs. Once TensorBoard is running, navigate your web browser to `localhost:6006` to view the TensorBoard.\n", "\n", "When looking at TensorBoard, you will see the navigation tabs in the top right corner. Each tab represents a set of serialized data that can be visualized." ] }, { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "## Visualizing Embeddings\n", "Visualising embeddings is a powerful technique! It helps you understand what your algorithm learned, and if this is what you expected it to learn.\n", "\n", "TensorBoard includes the Embedding Projector, a tool that lets you interactively visualize embeddings. This tool can read embeddings from your model and render them in two or three dimensions.\n", "\n", "The Embedding Projector has three panels:\n", "\n", "* Data panel on the top left, where you can choose the run, the embedding variable and data columns to color and label points by.\n", "* Projections panel on the bottom left, where you can choose the type of projection.\n", "* Inspector panel on the right side, where you can search for particular points and see a list of nearest neighbors.\n", "\n", "Here I'm particularly interested in projections. The Embedding Projector has three methods of reducing the dimensionality of a data set: two linear and one nonlinear. Each method can be used to create either a two- or three-dimensional view.\n", "\n", "**Principal Component Analysis** \n", "A straightforward technique for reducing dimensions is Principal Component Analysis (PCA). The Embedding Projector computes the top 10 principal components. The menu lets you project those components onto any combination of two or three. PCA is a linear projection, often effective at examining global geometry.\n", "\n", "![PCA](images/PCA-Vis.gif)\n", "\n", "**t-SNE** \n", "A popular non-linear dimensionality reduction technique is t-SNE. The Embedding Projector offers both two- and three-dimensional t-SNE views. Layout is performed client-side animating every step of the algorithm. Because t-SNE often preserves some local structure, it is useful for exploring local neighborhoods and finding clusters.\n", "\n", "![T-SNE](images/Tsne-Vis.png)\n", "\n", "**Custom** You can also construct specialized linear projections based on text searches for finding meaningful directions in space. To define a projection axis, enter two search strings or regular expressions. The program computes the centroids of the sets of points whose labels match these searches, and uses the difference vector between centroids as a projection axis.\n", "\n", "![Custom](images/Custom-Vis.png)\n", "\n", "## Further Exploration\n", "You can explore visually by zooming, rotating, and panning using natural click-and-drag gestures. Hovering your mouse over a point will show any metadata for that point. You can also inspect nearest-neighbor subsets. Clicking on a point causes the right pane to list the nearest neighbors, along with distances to the current point. The nearest-neighbor points are also highlighted in the projection. It is sometimes useful to restrict the view to a subset of points and perform projections only on those points. To do so, you can select points in multiple ways:\n", "\n", "* After clicking on a point, its nearest neighbors are also selected.\n", "* After a search, the points matching the query are selected.\n", "* Enabling selection, clicking on a point and dragging defines a selection sphere." ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.6.4" } }, "nbformat": 4, "nbformat_minor": 2 }