{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HFpLgHPVmtrt"
      },
      "source": [
        "<img src=\"https://cdn.comet.ml/img/notebook_logo.png\">"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L3wXps5XChOO"
      },
      "source": [
        "# Introduction\n",
        "\n",
        "[Artifacts](https://www.comet.com/site/products/artifacts-dataset-management/?utm_campaign-artifacts-launch&utm_source=colab-example&utm_medium=frontmatter) is a new tool that provides Machine Learning Teams with a convenient way to log, version, and access data from all parts of the experimentation pipeline. \n",
        "\n",
        "\n",
        "## What are Artifacts?\n",
        "\n",
        "We’ve built Comet Artifacts to help Machine Learning teams solve the challenges of iterating on datasets and tracking pipelines where the data generated from one experiment is fed into another experiment.\n",
        "\n",
        "An Artifact is composed of Artifact versions. Each Artifact has a name, a type, description, tags, and metadata.\n",
        "\n",
        "An Artifact version is a snapshot of files and assets, arranged in a folder-like logical structure. This snapshot can be tracked using metadata, a version number, tags, and aliases. A version tracks which experiments consumed it, and which experiment produced it.\n",
        "\n",
        "For a more complete overview [check out our full annoucement here](https://www.comet.com/site/blog/announcing-comet-artifacts/?utm_campaign-artifacts-launch&utm_source=colab-example&utm_medium=additional-resources)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xiyFPPFi6OrT"
      },
      "source": [
        "# Setup\n",
        "\n",
        "Install Comet and initialize a Project to try out Artifacts "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9xG1W2gayTck"
      },
      "outputs": [],
      "source": [
        "%pip install comet_ml pandas scikit-learn joblib"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-HXASUjXyW3M"
      },
      "outputs": [],
      "source": [
        "import comet_ml\n",
        "\n",
        "comet_ml.init(project_name=\"guide-artifacts-demo\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vLc_qnULNTS0"
      },
      "source": [
        "# Getting the Data\n",
        "\n",
        "For this example, we will use the California Housing Prices Dataset. Lets load the data and create a training and test set.  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VtsItkYN0t54"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "from sklearn.datasets import fetch_california_housing as load_data\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "dataset = load_data()\n",
        "X, y = dataset.data, dataset.target\n",
        "featurecols = dataset.feature_names\n",
        "\n",
        "# Train-Test Split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)\n",
        "\n",
        "train_df = pd.DataFrame(X_train, columns=featurecols)\n",
        "test_df = pd.DataFrame(X_test, columns=featurecols)\n",
        "\n",
        "train_df[\"target\"] = y_train\n",
        "test_df[\"target\"] = y_test"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-fDn9MNK1teM"
      },
      "outputs": [],
      "source": [
        "train_df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oAtIItHO1xec"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "\n",
        "os.makedirs(\"./datasets\", exist_ok=True)\n",
        "\n",
        "train_df.to_csv(\"./datasets/train.csv\", index=False)\n",
        "test_df.to_csv(\"./datasets/test.csv\", index=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y0qRHLQBNYpl"
      },
      "source": [
        "# Creating an Artifact to track the Dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LUyh2ZJ4Ur9y"
      },
      "source": [
        "Let's track our dataset with an Artifact. In order to create an Artifact, you will have to provide a name for it. You also have the option of providing additional information about the Artifact. You can provide a type string that identifies what kind of Artifact you are uploading (a model, dataset, etc). \n",
        "\n",
        "You can add alias identifiers to the Artifact, such as \"test data\" or \"staging model\". Later in this tutorial we will show you how Artifacts can be retrieved based on these aliases. \n",
        "\n",
        "Finally, you can attach a metadata dictionary to both the individual data assets uploaded to an Artifact as well as the Artifact itself. You can add any additional information about your Artifact in this dictionary.  \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "w7U5Q0Ou1-iP"
      },
      "outputs": [],
      "source": [
        "# Create a Comet Artifact\n",
        "artifact = comet_ml.Artifact(\n",
        "    name=\"california\",\n",
        "    artifact_type=\"dataset\",\n",
        "    aliases=[\"raw\"],\n",
        "    metadata={\"task\": \"regression\"},\n",
        ")\n",
        "\n",
        "# Add files to the Artifact\n",
        "for split, asset in zip(\n",
        "    [\"train\", \"test\"], [\"./datasets/train.csv\", \"./datasets/test.csv\"]\n",
        "):\n",
        "    artifact.add(asset, metadata={\"dataset_stage\": \"raw\", \"dataset_split\": split})\n",
        "\n",
        "experiment = comet_ml.Experiment()\n",
        "experiment.add_tag(\"upload\")\n",
        "experiment.log_artifact(artifact)\n",
        "\n",
        "experiment.end()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K9l11MP2Y5W3"
      },
      "source": [
        "In your Workspace, you will see an Artifacts tab where you can view the data that has been uploaded. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YeOo1D4TezCl"
      },
      "source": [
        "![Screenshot 2023-01-10 at 23-15-56 Comet.ml - Supercharging Machine Learning.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lKlQUvmmZscF"
      },
      "source": [
        "Clicking on the Artifact will bring up the Version information and associated Metadata. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0_4q8B8ee_Ez"
      },
      "source": [
        "![Screenshot 2023-01-10 at 23-16-23 Comet.ml - Supercharging Machine Learning.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DI34eveoNtuc"
      },
      "source": [
        "# Using an Artifact"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G7YJYBrxNyj9"
      },
      "source": [
        "Now that we have an Artifact tracking our dataset, let's move on to using it to train a model!  \n",
        "\n",
        "\n",
        "First, let's make a directory to save our Artifacts."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "p-DvX6Pe2e0p"
      },
      "outputs": [],
      "source": [
        "! mkdir ./artifacts"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3f47rvl9vPz4"
      },
      "source": [
        "### Download the Artifact"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d0QtzIuzaOvL"
      },
      "source": [
        "We can fetch the Artifact we need using its name, and either the version or alias.  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "89PLtHw53dX8"
      },
      "outputs": [],
      "source": [
        "experiment = comet_ml.Experiment()\n",
        "experiment.add_tag(\"train\")\n",
        "\n",
        "# Fetch the Artifact object from Comet\n",
        "name = \"california\"\n",
        "version_or_alias = \"raw\"\n",
        "artifact = experiment.get_artifact(name, version_or_alias=version_or_alias)\n",
        "\n",
        "# Download Artifact\n",
        "output_path = \"./artifacts\"\n",
        "artifact.download(output_path, overwrite_strategy=\"PRESERVE\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vDr2VsVSRKfN"
      },
      "source": [
        "### Train a Model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UYLYqoMH6H08"
      },
      "outputs": [],
      "source": [
        "from sklearn.linear_model import LinearRegression\n",
        "from joblib import dump\n",
        "\n",
        "# Load Data from Artifact\n",
        "train_df = pd.read_csv(\"./artifacts/train.csv\")\n",
        "test_df = pd.read_csv(\"./artifacts/test.csv\")\n",
        "\n",
        "y_train = train_df.pop(\"target\").values\n",
        "X_train = train_df.values\n",
        "\n",
        "y_test = test_df.pop(\"target\").values\n",
        "X_test = test_df.values\n",
        "\n",
        "# Initialize Model\n",
        "model = LinearRegression()\n",
        "model.fit(X_train, y_train)\n",
        "\n",
        "# Evaluate Model\n",
        "train_score = model.score(X_train, y_train)\n",
        "test_score = model.score(X_test, y_test)\n",
        "\n",
        "experiment.log_metric(\"train-score\", train_score)\n",
        "experiment.log_metric(\"test-score\", test_score)\n",
        "\n",
        "# Save Model\n",
        "model_path = \"./linear-model.pkl\"\n",
        "dump(model, model_path)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Whe6tZBNRV7R"
      },
      "source": [
        "# Log Model as an Artifact\n",
        "\n",
        "Let's log the model we just trained as an Artifact. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GBL50p59RUui"
      },
      "outputs": [],
      "source": [
        "# Log Model as an Artifact\n",
        "model_artifact = comet_ml.Artifact(\n",
        "    \"housing-model\", artifact_type=\"model\", aliases=[\"baseline\"]\n",
        ")\n",
        "model_artifact.add(model_path)\n",
        "experiment.log_artifact(model_artifact)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0_M1HZ8_dH-S"
      },
      "source": [
        "You can view the Artifacts Produced and Consumed by an Experiment in the \"Assets and Artifacts\" tab under Artifacts. Toggle the direction selector to filter by Input, which refers to Artifacts that were consumed, and Output which refers to Artifacts that were produced  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "flCC_reqcFbF"
      },
      "outputs": [],
      "source": [
        "experiment.display(tab=\"assets\")\n",
        "experiment.end()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gQHLMYXMOCPq"
      },
      "source": [
        "# Updating an Artifact "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6ablGuV8bNgV"
      },
      "source": [
        "Our scores on the raw dataset were not that great. Why don't we scale the data and update our Artifact to reflect this.  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "t1RtWsxy7uec"
      },
      "outputs": [],
      "source": [
        "# Scores aren't great, so lets scale the features\n",
        "from sklearn.preprocessing import StandardScaler as Scaler\n",
        "\n",
        "experiment = comet_ml.Experiment()\n",
        "experiment.add_tag(\"upload\")\n",
        "\n",
        "X_scaler = Scaler().fit(X_train)\n",
        "y_scaler = Scaler().fit(y_train.reshape(-1, 1))\n",
        "\n",
        "X_train_scaled = X_scaler.transform(X_train)\n",
        "X_test_scaled = X_scaler.transform(X_test)\n",
        "\n",
        "y_train_scaled = y_scaler.transform(y_train.reshape(-1, 1))\n",
        "y_test_scaled = y_scaler.transform(y_test.reshape(-1, 1))\n",
        "\n",
        "train_scaled_df = pd.DataFrame(X_train, columns=featurecols)\n",
        "test_scaled_df = pd.DataFrame(X_test, columns=featurecols)\n",
        "\n",
        "train_scaled_df[\"target\"] = y_train\n",
        "test_scaled_df[\"target\"] = y_test\n",
        "\n",
        "train_scaled_df.to_csv(\"./datasets/train-scaled.csv\")\n",
        "test_scaled_df.to_csv(\"./datasets/test-scaled.csv\")\n",
        "\n",
        "# Update Artifact with Scaled Data\n",
        "scaled_dataset_artifact = comet_ml.Artifact(\n",
        "    \"california\",\n",
        "    artifact_type=\"dataset\",\n",
        "    aliases=[\"standard-scaled\"],\n",
        "    metadata={\"task\": \"regression\"},\n",
        ")\n",
        "# Add files to the Artifact\n",
        "for split, asset in zip(\n",
        "    [\"train\", \"test\"], [\"./datasets/train-scaled.csv\", \"./datasets/test-scaled.csv\"]\n",
        "):\n",
        "    scaled_dataset_artifact.add(\n",
        "        asset, metadata={\"dataset_stage\": \"standard-scaled\", \"dataset_split\": split}\n",
        "    )\n",
        "\n",
        "experiment.log_artifact(scaled_dataset_artifact)\n",
        "experiment.end()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zt54EPm6OYlV"
      },
      "source": [
        "# Train a Model with the Latest Version of the Dataset Artifact"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9UuAGSYEFS7L"
      },
      "outputs": [],
      "source": [
        "experiment = comet_ml.Experiment()\n",
        "experiment.add_tag(\"train\")\n",
        "\n",
        "# Fetch the Artifact object from Comet\n",
        "name = \"california\"\n",
        "version_or_alias = \"standard-scaled\"\n",
        "artifact = experiment.get_artifact(name, version_or_alias=version_or_alias)\n",
        "\n",
        "\n",
        "# Download Artifact\n",
        "output_path = \"./artifacts\"\n",
        "artifact.download(output_path, overwrite_strategy=\"PRESERVE\")\n",
        "\n",
        "# Load Data from Artifact\n",
        "train_df = pd.read_csv(\"./artifacts/train-scaled.csv\")\n",
        "test_df = pd.read_csv(\"./artifacts/test-scaled.csv\")\n",
        "\n",
        "y_train = train_df.pop(\"target\").values\n",
        "X_train = train_df.values\n",
        "\n",
        "y_test = test_df.pop(\"target\").values\n",
        "X_test = test_df.values\n",
        "\n",
        "# Initialize Model\n",
        "model = LinearRegression()\n",
        "model.fit(X_train, y_train)\n",
        "\n",
        "# Evaluate Model\n",
        "train_score = model.score(X_train, y_train)\n",
        "test_score = model.score(X_test, y_test)\n",
        "\n",
        "experiment.log_metric(\"train-score\", train_score)\n",
        "experiment.log_metric(\"test-score\", test_score)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "t2Vc-NxGdFhK"
      },
      "outputs": [],
      "source": [
        "experiment.end()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_HQe28iuj5sn"
      },
      "source": [
        "Doesn't look like the scaling helped :( Back to the drawing board!"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k-DU-t_xiJ_x"
      },
      "source": [
        "# Conclusion\n",
        "\n",
        "We hope you enjoyed this introductory guide to Artifacts, a simple, light weight way to version your datasets and models, while providing information about the lineage of your data through your experiments. \n",
        "\n",
        "Interested in learning more about Artifacts? Check out the [docs](https://www.comet.ml/docs/user-interface/artifacts/?utm_campaign-artifacts-launch&utm_source=colab-example&utm_medium=additional-resources)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "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": "3.9.1"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}