library_tour.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "library_tour.ipynb",
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyNGoc71l3lquzf4+kbbNLey",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ContextLab/psyc32-library-tour/blob/main/library_tour.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jDhIyr6XsghX"
      },
      "source": [
        "# The *Library Tour*: Introduction and Overview\n",
        "\n",
        "![Library GIF](https://media1.giphy.com/media/3o7WIu0DcVJxQbNWb6/giphy.gif)\n",
        "\n",
        "This assignment is intended to give you practice using several popular Python libraries used in data science, research, analysis, and other computational fields.  You'll do a series of sample problems that utilize the following libraries:\n",
        "- [NumPy](https://numpy.org/): efficient matrix math library\n",
        "- [Pandas](https://pandas.pydata.org/): data organization and manipulation\n",
        "- [SciPy](https://www.scipy.org/): high-level mathematical operations\n",
        "- [Scikit-Learn](https://scikit-learn.org/stable/): machine learning\n",
        "- [HyperTools](hypertools.readthedocs.io/): high-level data visualization and manipulation\n",
        "- [Matplotlib](https://matplotlib.org/) and [seaborn](https://seaborn.pydata.org/): data visualization\n",
        "\n",
        "We'll be analyzing two classic datasets:\n",
        "- [MNIST](http://yann.lecun.com/exdb/mnist/): a set of 70K images of handwritten digits, with \"ground truth\" labels\n",
        "- [Murdock (1926)](https://www.dropbox.com/s/3wyy2b2xlqygokf/Murd62.pdf): a [free recall](https://en.wikipedia.org/wiki/Free_recall) dataset reported in one of the first studies to identify the [primacy](https://en.wikipedia.org/wiki/Primacy_effect) and [recency](https://en.wikipedia.org/wiki/Recency_effect) effects.\n",
        "\n",
        "Follow the instructions in the cells below to complete your library tour!\n",
        "\n",
        "## Grading\n",
        "\n",
        "This assignment is worth a total of 15 points.  You may view the tests in the public rubric [here](https://github.com/ContextLab/cs-for-psych/blob/master/assignments/library-tour/public_rubric.xls).  To get full credit, you must (correctly) fill in all of the code between the commented `### BEGIN YOUR CODE` and `### END YOUR CODE` blocks throughout this notebook.  (No other code should be modified.)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HU1H4uHYsXLB"
      },
      "source": [
        "# Library imports-- take note of these aliases!\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import sklearn as skl\n",
        "import matplotlib as mpl\n",
        "import seaborn as sns\n",
        "import warnings\n",
        "warnings.simplefilter('ignore')\n",
        "\n",
        "try:\n",
        "  import hypertools as hyp\n",
        "except ModuleNotFoundError:\n",
        "  !pip install --quiet git+https://github.com/ContextLab/hypertools.git\n",
        "  import hypertools as hyp"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ou0TkS-1mtDz"
      },
      "source": [
        "# Load the datasets and wrangle them into convenient formats\n",
        "\n",
        "Although you won't need to fill in any code in the cells below that load and wrangle the datasets, you should make sure that you understand the code and how each dataset is formatted."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hJP9lc5OCs-Z"
      },
      "source": [
        "## Modified National Institute of Standards and Technology (MNIST)\n",
        "\n",
        "We'll download the [MNIST dataset](https://en.wikipedia.org/wiki/MNIST_database) from [OpenML](https://www.openml.org/s/88/data) using Scikit-learn's `fetch_openml` function.  We'll store the observations (images of each hand-drawn digit) in rows of a Pandas DataFrame, and the labels (0 -- 9) in a Pandas Series object."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LR7J-5uvzqZe"
      },
      "source": [
        "# Load the datasets into a convenient format\n",
        "from sklearn.datasets import fetch_openml\n",
        "\n",
        "#MNIST\n",
        "try:\n",
        "  mnist_obs, mnist_labels = fetch_openml('mnist_784', version=1, return_X_y=True, as_frame=True)\n",
        "except: #OpenML sometimes throws an error; use a different source if needed\n",
        "  import tensorflow as tf\n",
        "  mnist = tf.keras.datasets.mnist.load_data(path='mnist.npz')\n",
        "  mnist_obs = np.concatenate([mnist[0][0], mnist[1][0]], axis=0)\n",
        "  mnist_labels = np.concatenate([mnist[0][1], mnist[1][1]], axis=0)\n",
        "  mnist_labels = np.array([str(x) for x in mnist_labels])\n",
        "\n",
        "  mnist_obs = mnist_obs.reshape([mnist_obs.shape[0], mnist_obs.shape[1] * mnist_obs.shape[2]]) / 255.0\n",
        "  mnist_obs = pd.DataFrame(data=mnist_obs, columns=[f'pixel{i}' for i in range(1, mnist_obs.shape[1] + 1)])\n",
        "  del mnist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BHFc4c7hndHI"
      },
      "source": [
        "Now let's explore what the MNIST data look like.  We can plot the first ten rows of the `mnist_obs` DataFrame to peek at some raw observations:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 412
        },
        "id": "XGrdYqEwn8zg",
        "outputId": "5a3c47a5-bad8-4d69-afb7-fd0e19f5b476"
      },
      "source": [
        "mnist_obs.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>pixel1</th>\n",
              "      <th>pixel2</th>\n",
              "      <th>pixel3</th>\n",
              "      <th>pixel4</th>\n",
              "      <th>pixel5</th>\n",
              "      <th>pixel6</th>\n",
              "      <th>pixel7</th>\n",
              "      <th>pixel8</th>\n",
              "      <th>pixel9</th>\n",
              "      <th>pixel10</th>\n",
              "      <th>pixel11</th>\n",
              "      <th>pixel12</th>\n",
              "      <th>pixel13</th>\n",
              "      <th>pixel14</th>\n",
              "      <th>pixel15</th>\n",
              "      <th>pixel16</th>\n",
              "      <th>pixel17</th>\n",
              "      <th>pixel18</th>\n",
              "      <th>pixel19</th>\n",
              "      <th>pixel20</th>\n",
              "      <th>pixel21</th>\n",
              "      <th>pixel22</th>\n",
              "      <th>pixel23</th>\n",
              "      <th>pixel24</th>\n",
              "      <th>pixel25</th>\n",
              "      <th>pixel26</th>\n",
              "      <th>pixel27</th>\n",
              "      <th>pixel28</th>\n",
              "      <th>pixel29</th>\n",
              "      <th>pixel30</th>\n",
              "      <th>pixel31</th>\n",
              "      <th>pixel32</th>\n",
              "      <th>pixel33</th>\n",
              "      <th>pixel34</th>\n",
              "      <th>pixel35</th>\n",
              "      <th>pixel36</th>\n",
              "      <th>pixel37</th>\n",
              "      <th>pixel38</th>\n",
              "      <th>pixel39</th>\n",
              "      <th>pixel40</th>\n",
              "      <th>...</th>\n",
              "      <th>pixel745</th>\n",
              "      <th>pixel746</th>\n",
              "      <th>pixel747</th>\n",
              "      <th>pixel748</th>\n",
              "      <th>pixel749</th>\n",
              "      <th>pixel750</th>\n",
              "      <th>pixel751</th>\n",
              "      <th>pixel752</th>\n",
              "      <th>pixel753</th>\n",
              "      <th>pixel754</th>\n",
              "      <th>pixel755</th>\n",
              "      <th>pixel756</th>\n",
              "      <th>pixel757</th>\n",
              "      <th>pixel758</th>\n",
              "      <th>pixel759</th>\n",
              "      <th>pixel760</th>\n",
              "      <th>pixel761</th>\n",
              "      <th>pixel762</th>\n",
              "      <th>pixel763</th>\n",
              "      <th>pixel764</th>\n",
              "      <th>pixel765</th>\n",
              "      <th>pixel766</th>\n",
              "      <th>pixel767</th>\n",
              "      <th>pixel768</th>\n",
              "      <th>pixel769</th>\n",
              "      <th>pixel770</th>\n",
              "      <th>pixel771</th>\n",
              "      <th>pixel772</th>\n",
              "      <th>pixel773</th>\n",
              "      <th>pixel774</th>\n",
              "      <th>pixel775</th>\n",
              "      <th>pixel776</th>\n",
              "      <th>pixel777</th>\n",
              "      <th>pixel778</th>\n",
              "      <th>pixel779</th>\n",
              "      <th>pixel780</th>\n",
              "      <th>pixel781</th>\n",
              "      <th>pixel782</th>\n",
              "      <th>pixel783</th>\n",
              "      <th>pixel784</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>147.0</td>\n",
              "      <td>252.0</td>\n",
              "      <td>42.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10 rows × 784 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "   pixel1  pixel2  pixel3  pixel4  ...  pixel781  pixel782  pixel783  pixel784\n",
              "0     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "1     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "2     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "3     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "4     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "5     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "6     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "7     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "8     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "9     0.0     0.0     0.0     0.0  ...       0.0       0.0       0.0       0.0\n",
              "\n",
              "[10 rows x 784 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5cr4B-wmoNgo"
      },
      "source": [
        "We can plot individual images by re-shaping them into 28 x 28 images (e.g., using Numpy's `reshape` function), and then displaying the result (e.g., using seaborn's `heatmap` function):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 822
        },
        "id": "8S-r-7fjojHX",
        "outputId": "b1c06269-e135-474d-a83b-58b10db5c80c"
      },
      "source": [
        "def plot_digit_image(img):\n",
        "  sns.heatmap(np.array(img).reshape([28, 28]), cmap='gray', cbar=False)\n",
        "\n",
        "#plot the first few images along with their labels\n",
        "for i in range(3):\n",
        "  print(f'The next image\\'s label is {mnist_labels[i]}:')\n",
        "  plot_digit_image(mnist_obs.loc[i])\n",
        "  mpl.pyplot.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The next image's label is 5:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "The next image's label is 0:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "The next image's label is 4:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZdgaTO6e8lqv"
      },
      "source": [
        "## Murdock (1962) Free Recall dataset\n",
        "\n",
        "The Murdock (1962) free recall dataset is organized into a tar file with the following structure (non-essential files and folders are omitted):\n",
        "\n",
        "```\n",
        "Murd62:\n",
        "├── fr10-2.txt\n",
        "├── fr15-2.txt\n",
        "├── fr20-1.txt\n",
        "├── fr20-2.txt\n",
        "├── fr30-1.txt\n",
        "└── fr40-1.txt\n",
        "```\n",
        "\n",
        "In the experiment, six groups of participants studied random word lists and then wrote down as many words as they could remember, in any order they wished.  Across different experimental conditions, Murdock systematically varied the list length (i.e., the number of words participants in that condition studied on each list) and the presentation rate.\n",
        "\n",
        "Each `.txt` file comprises data from one experimental condition.  The first number in each filename (after `fr`) denotes the list length and the second number in each filename (after the hyphen) denotes the presentation rate, in Hz.  For example, the `fr10-2.txt` file includes data from participants who studied 10-word lists presented at a rate of 2 words per second.  Each row of each file corresponds to one participant/trial.  For example, if a row contains the sequence:\n",
        "\n",
        "```\n",
        "20 19 18 15 1 5 8\n",
        "```\n",
        "\n",
        "then that means that the participant wrote down the 20th-presented word first, followed by the 19th word, and so on.  The code `88` means that the participant wrote down a word that had not appeared on the list they had studied.  Any other invalid entries (i.e., values other than 88 that are greater than the list length, or values that are less than or equal to 0) should be ignored."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oj-evn3xvKfo"
      },
      "source": [
        "import os\n",
        "import tarfile\n",
        "import urllib.request as get\n",
        "from glob import glob as lsdir\n",
        "\n",
        "# fetch and extract a tarfile stored remotely\n",
        "def extract_remote_tarfile(url):\n",
        "  stream = get.urlopen(url)\n",
        "  tf = tarfile.open(fileobj=stream, mode='r|gz')\n",
        "  tf.extractall()\n",
        "\n",
        "# load a text file\n",
        "def txt_loader(fname):\n",
        "  with open(fname, 'r') as f:\n",
        "    return f.read()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V7P_-wnplLE8"
      },
      "source": [
        "murdock_url = 'http://memory.psych.upenn.edu/files/pubs/Murd62.data.tgz'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4NXg15q9llzx"
      },
      "source": [
        "extract_remote_tarfile(murdock_url)\n",
        "murdock_root = 'Murd62'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "quGGhvchBFz-"
      },
      "source": [
        "def load_murdock_file(fname):\n",
        "  data = txt_loader(fname).split('\\n')\n",
        "  \n",
        "  #create a list of lists from the data\n",
        "  recalls = [[int(x) if ((int(x) > 0) or (int(x) == 88)) else np.nan for x in d.strip().split(' ')] for d in data if len(d.strip()) > 0]\n",
        "  df = pd.DataFrame(recalls)\n",
        "  df.index.rename('List', inplace=True)\n",
        "  df.columns.name = 'Recall'\n",
        "  return df"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bMS_J-WdHpgU"
      },
      "source": [
        "murdock_data = {os.path.split(f)[1][2:-4]: load_murdock_file(f) for f in lsdir(os.path.join(murdock_root, 'fr*.txt'))}"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "04HwQXE5Kw_F"
      },
      "source": [
        "The dictionary `murdock_data` has keys corresponding to the experimental conditions ('10-2', '15-2', '20-1', '20-2', '30-1', and '40-1').  Each value in the dictionary is a `DataFrame` whose rows each reflect a single participant's recalls for a single list, and whose columns reflect individual responses (recalls).  Note that the total number of columns in each condition's DataFrame is equal to the *maximum* number of responses across all lists and participants, for that condition.  Any lists that received fewer responses will be padded with `NaN`s.  Some rows may also contain *all* `NaN`s if there were no valid responses for that list.\n",
        "\n",
        "As an example, let's take a look at the recall data for the first lists from the `20-2` condition:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "ufnMGydYL7MD",
        "outputId": "76bcf28e-e5d7-41e2-ce9b-6f31a296ea7b"
      },
      "source": [
        "murdock_data['20-2'].head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th>Recall</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "      <th>8</th>\n",
              "      <th>9</th>\n",
              "      <th>10</th>\n",
              "      <th>11</th>\n",
              "      <th>12</th>\n",
              "      <th>13</th>\n",
              "      <th>14</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>List</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>12</td>\n",
              "      <td>7</td>\n",
              "      <td>10</td>\n",
              "      <td>20.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>19</td>\n",
              "      <td>20</td>\n",
              "      <td>18</td>\n",
              "      <td>10.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>88.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>18</td>\n",
              "      <td>19</td>\n",
              "      <td>20</td>\n",
              "      <td>9.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>88.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>19</td>\n",
              "      <td>88</td>\n",
              "      <td>17</td>\n",
              "      <td>10.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>19</td>\n",
              "      <td>20</td>\n",
              "      <td>17</td>\n",
              "      <td>18.0</td>\n",
              "      <td>88.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "Recall  0   1   2     3     4     5     6     7   8   9   10  11  12  13  14\n",
              "List                                                                        \n",
              "0       12   7  10  20.0  15.0  19.0   6.0  17.0 NaN NaN NaN NaN NaN NaN NaN\n",
              "1       19  20  18  10.0  14.0  11.0  15.0  88.0 NaN NaN NaN NaN NaN NaN NaN\n",
              "2       18  19  20   9.0  11.0  88.0   NaN   NaN NaN NaN NaN NaN NaN NaN NaN\n",
              "3       19  88  17  10.0   6.0   5.0   NaN   NaN NaN NaN NaN NaN NaN NaN NaN\n",
              "4       19  20  17  18.0  88.0   NaN   NaN   NaN NaN NaN NaN NaN NaN NaN NaN"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1tDClFyBMBQm"
      },
      "source": [
        "# Matrix operations and data wrangling with NumPy and Pandas"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CKoHGO0WtZRl"
      },
      "source": [
        "## Example digits\n",
        "\n",
        "In the MNIST dataset, what are the similarities and differences between images of the same digit?  Let's explore this by creating two functions:\n",
        "  - `average_digit` should take in `mnist_obs` (`obs`), `mnist_labels` (`labels`), and an integer between 0 and 9, inclusive (`digit`).  The function should return a 28 by 28 NumPy array containing the average image for the given digit.\n",
        "  - `std_digit` should take in `mnist_obs` (`obs`), `mnist_labels` (`labels`), and an integer between 0 and 9, inclusive (`digit`).  The function should return a 28 by 28 NumPy array containing the standard deviation (across images) for each pixel, for the given digit."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7UtgnpvzuGMY"
      },
      "source": [
        "def average_digit(obs, labels, digit):\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE\n",
        "\n",
        "def std_digit(obs, labels, digit):\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vSK9FiGYv1_5"
      },
      "source": [
        "You can check your functions by attempting to produce the heatmaps below:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 268
        },
        "id": "VzknrwqIuKSn",
        "outputId": "b7e8cf64-8d1d-47b1-e246-cd3d1626bfb4"
      },
      "source": [
        "sns.heatmap(average_digit(mnist_obs, mnist_labels, 0));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "c04YUeuavd56",
        "outputId": "9cd00086-f1cb-4a2c-9be6-c8b4ef93d0e7"
      },
      "source": [
        "sns.heatmap(std_digit(mnist_obs, mnist_labels, 0));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KYfwdZ04h_xi"
      },
      "source": [
        "Now let's create a mosaic of the averages and standard deviations across images:\n",
        "- The mosaic should be returned as a single NumPy array\n",
        "- Every block of 28 x 28 pixels should reflect data for one digit\n",
        "- The *columns* of the mosaic should reflect the digits 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9 (in that order)\n",
        "- The top *row* of the mosaic should show the mean images for each digit\n",
        "- The bottom *row* of the mosaic should show the pixel-wise standard deviations (across images) for each digit"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7V04oV-miuf_"
      },
      "source": [
        "def mosaic(obs, labels, rows=[average_digit, std_digit], columns=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]):\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HJRjdNEDlKXz"
      },
      "source": [
        "Now let's display the full mosaic as a heatmap:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xQpaD-0QlXkV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 194
        },
        "outputId": "c76fa113-d9d7-408f-e870-689c918c4587"
      },
      "source": [
        "fig, ax = mpl.pyplot.subplots(figsize=[15, 3]);\n",
        "sns.heatmap(mosaic(mnist_obs, mnist_labels), ax=ax, cbar=False);\n",
        "ax.get_xaxis().set_visible(False);\n",
        "ax.get_yaxis().set_visible(False);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wBox4myBtPZ_"
      },
      "source": [
        "## Serial position curves\n",
        "\n",
        "A *serial position curve* computes the average probability of recalling each item on a studied list as a function of its presentation position.\n",
        "\n",
        "Fill in the contents of the `spc` function skeleton below.\n",
        "- Your function should be able take in any DataFrame in the same format as those in the Murdock (1962) dataset.  You should return a DataFrame with a single row, whose columns are list presentation positions.  Optionally, the user may pass in an experimental condition string (`cond`) denoting the list length, then a hyphen, and then the presentation rate (e.g. `'10-2'`).\n",
        "- When passed multiple DataFrames in a dictionary, you should return a DataFrame with one row per inputted DataFrame, where the columns denote presentation positions and the rows are indexed by the dictionary's keys. (Hint: set `cond=k` for `k` in `dict.keys()`.)\n",
        "- Any entries with no data should be denoted as `NaN`s.\n",
        "\n",
        "Implementation notes:\n",
        "- The 10-2 condition's data file has an extra row in it.  This error is part of the original Murdock (1962) dataset.  When the `load_murdock_file` function loads in the dataset, the dataframe for the 10-2 condition will have an extra row of `NaN`s in it-- just treat that row as a list where no words were recalled.\n",
        "- In the original Murdock (1962) serial position curves below, you'll notice that the list positions begin with 1 and end with postion equal to the list length.  In your implementation, you should use 0-indexing (the first serial position should be 0 and the last should be list length minus 1)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wDM3IgHUM751"
      },
      "source": [
        "def spc(df, cond=''):\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "52cGtUiRVWoA"
      },
      "source": [
        "#compute serial position curves\n",
        "spcs = spc(murdock_data)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W07VRAWielx5"
      },
      "source": [
        "Plot the serial position curves (compare with Figure 1 from [Murdock (1962)](https://www.dropbox.com/s/3wyy2b2xlqygokf/Murd62.pdf)):\n",
        "\n",
        "![Serial position curves](https://raw.githubusercontent.com/ContextLab/cs-for-psych/master/assignments/library-tour/figs/Murd62_SPCs.png)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "HeC8KC_2bIQY",
        "outputId": "d427169b-e1d5-4fec-a703-29e597c4898e"
      },
      "source": [
        "spcs.T.plot(xlabel='Serial position', ylabel='Probability of recall');"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qmD4wfwAM_OL"
      },
      "source": [
        "# Exploratory analyses with HyperTools and SciPy\n",
        "\n",
        "A good \"first step\" towards understanding many datasets is to examine how the observations are related.  The HyperTools library provides some convenient ways of visualizing high-dimensional data.  Let's start by using HyperTools to explore the MNIST dataset.  We'll:\n",
        "  - Convert each digit image into a high-dimensional feature vector\n",
        "  - Use [Uniform Manifold Approximation and Projection (UMAP)](https://arxiv.org/pdf/1802.03426.pdf) to project the feature vectors onto a 2D plane\n",
        "  - Plot each image's (reduced) feature vector as a single point\n",
        "  - Color each point according to that image's ground truth label\n",
        "\n",
        "In the example below, we'll start by defining a dictionary of UMAP parameters (this step is not strictly necessary, but here I've chosen a set of parameters that I know will produce a \"nice looking\" projection).  Then we'll use the `hyp.plot` function to visualize the dataset (see the [API specification](https://hypertools.readthedocs.io/en/latest/api.html) for details about each parameter):\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 729
        },
        "id": "-lnuPbxzxz02",
        "outputId": "82590930-9ead-4327-c24d-917bdefa93f7"
      },
      "source": [
        "import random\n",
        "random.seed(a=132) #this will guarantee consistency across different runs of the notebook\n",
        "\n",
        "params = {'n_neighbors': 5, 'min_dist': 0.3, 'metric': 'correlation'}\n",
        "hyp.plot(mnist_obs, '.', reduce={'model': 'UMAP', 'params': params}, hue=np.array(mnist_labels), ndims=2, size=[10, 10]);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fqn_oJL76gNh"
      },
      "source": [
        "The fact that we can see 10 distinct clusters of dots, each with distinct (and largely uniform) colors, means that the digit images are (mostly) separable.  In other words, images of the same digit tend to be similar to each other, and images of different digits tend to be different from each other.  When we see well-defined clusters like in this example, it is a good indication that classifiers will be able to distinguish between the clusters.  We could gain additional insights by asking other questions about the data that are driven by the above visualization.  For example:\n",
        "  - Which clusters (or data labels) are spread out versus compact?  More compact clusters indicate greater uniformity.\n",
        "  - Which clusters show intermixing?  This indicates potential for confusion across clusters, or potential relations across labels or conditions.\n",
        "  - Which clusters are nearby versus distant?  This indicates which observations are similar versus different from each other."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wbkSHsTkNHHQ"
      },
      "source": [
        "# Machine learning with Scikit-Learn\n",
        "\n",
        "Let's learn about scikit-learn by training a [logistic regression](https://en.wikipedia.org/wiki/Logistic_regression) classifier to identity the digit labels.\n",
        "\n",
        "First, we'll use the `skl.model_selection.train_test_split` function to divide the dataset into a training set (which we'll use to train our classifier-- i.e., to fit the model's parameters) and a testing set (which we'll use to evaluate the classifier's performance."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ILp5_XmC-gTT",
        "outputId": "ddd2b500-dd51-479a-ade6-9752d2ac7f52"
      },
      "source": [
        "mnist_train_obs, mnist_test_obs, mnist_train_labels, mnist_test_labels = skl.model_selection.train_test_split(mnist_obs, mnist_labels, test_size=0.25, random_state=10)\n",
        "\n",
        "#print out some details of the training and test sets\n",
        "print(f'The original dataset has {len(mnist_labels)} observations.')\n",
        "print(f'After splitting, the training set has {len(mnist_train_labels)} observations and the test set has {len(mnist_test_labels)} observations.')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The original dataset has 70000 observations.\n",
            "After splitting, the training set has 52500 observations and the test set has 17500 observations.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PfjpSoErAxxX"
      },
      "source": [
        "Next, let's set up the classifier and fit it to the training data (we'll stick with the default classifier options and parameters):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Lv43hQD1BE32",
        "outputId": "42467010-bcd9-4e6c-b6cd-43facfd01ad7"
      },
      "source": [
        "classifier = skl.linear_model.LogisticRegression()\n",
        "classifier.fit(mnist_train_obs, mnist_train_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rj2FvCSIBcWu"
      },
      "source": [
        "And now that the classifier has been trained, we can feed in new images that the classifier hasn't \"seen\" yet and get predicted labels:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_7MvBdS6Bngm"
      },
      "source": [
        "predicted_test_labels = classifier.predict(mnist_test_obs)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bn9M43mMJJhI"
      },
      "source": [
        "Finally, let's print out a report summarizing how the classifier performed, along with a [confusion matrix](https://en.wikipedia.org/wiki/Confusion_matrix):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 626
        },
        "id": "QPKOCs_N_uAi",
        "outputId": "c131309a-2735-424f-bceb-e9aac7e478e9"
      },
      "source": [
        "print(skl.metrics.classification_report(mnist_test_labels, predicted_test_labels))\n",
        "\n",
        "print('\\n\\nConfusion matrix:')\n",
        "skl.metrics.ConfusionMatrixDisplay.from_estimator(classifier, mnist_test_obs, mnist_test_labels);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.97      0.96      1757\n",
            "           1       0.96      0.98      0.97      1901\n",
            "           2       0.91      0.90      0.90      1784\n",
            "           3       0.89      0.91      0.90      1804\n",
            "           4       0.94      0.92      0.93      1708\n",
            "           5       0.89      0.85      0.87      1566\n",
            "           6       0.93      0.96      0.94      1657\n",
            "           7       0.94      0.92      0.93      1873\n",
            "           8       0.88      0.87      0.87      1720\n",
            "           9       0.89      0.92      0.91      1730\n",
            "\n",
            "    accuracy                           0.92     17500\n",
            "   macro avg       0.92      0.92      0.92     17500\n",
            "weighted avg       0.92      0.92      0.92     17500\n",
            "\n",
            "\n",
            "\n",
            "Confusion matrix:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JYujryJuK72R"
      },
      "source": [
        "We've achieved 92% accuracy (accuracy = 0.92)-- not bad!\n",
        "\n",
        "In the example above, we used the keyword argument `test_size=0.25` (passed to `train_test_split`) to specify that 75% of the dataset should be used for training and 25% should be used for testing.  How does our classifier's performance (on the test set) vary as a function of the training versus test set size?\n",
        "\n",
        "Write a function to compute the classification accuracy as a function of test size.  Your function should:\n",
        "\n",
        "1. Take in a list of test sizes (either floats to specify proportions, or numbers of samples)\n",
        "2. For each test size in turn, use `train_test_split` (specifying `random_state=10`) to divide the dataset into training and test sets\n",
        "3. Train a logisitic regression classfier (with all of its default arguments) on the training data\n",
        "4. Predict the labels for the test data and compute the accuracy (accuracy is the number of correctly labeled test examples divided by the total number of test examples).  Note: this should be a float between 0 and 1, inclusive.\n",
        "5. Return a list of the resulting accuracies (in the same order as the test size list that you passed into the function)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y6h6zvmOAlE3"
      },
      "source": [
        "def accuracy_by_test_size(data, labels, sizes):\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gcCWFfmdRBg8"
      },
      "source": [
        "Example use and figure to help with debugging (think about how you could use a plot like this to help you decide how to divide up your data into training versus test sets!):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "-bcssyIROweN",
        "outputId": "3c3cae3e-c02d-4d14-839f-7e3d09fe002c"
      },
      "source": [
        "sizes = [0.99, 0.95, 0.9, 0.75, 0.5, 0.25, 0.1, 0.05, 0.01]\n",
        "accuracies = accuracy_by_test_size(mnist_obs.iloc[:1000], mnist_labels[:1000], sizes)\n",
        "\n",
        "mpl.pyplot.plot(1 - np.array(sizes), accuracies, color='k');\n",
        "mpl.pyplot.xlabel('Training set size (proportion of observations)');\n",
        "mpl.pyplot.ylabel('Classification accuracy');"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OPSfxMeDNPAa"
      },
      "source": [
        "# Data visualization with Matplotlib and Seaborn\n",
        "\n",
        "On the last stop on this tour, your job is to create a bar plot using `mpl.pyplot.bar`.  You should match the example figure below as closely as possible to get full credit:\n",
        "\n",
        "![reference figure](https://raw.githubusercontent.com/ContextLab/cs-for-psych/master/assignments/library-tour/figs/accuracy_by_condition_reference.png)\n",
        "\n",
        "Here are the details:\n",
        "- Sort the conditions in the Murdock (1962) dataset in ascending order of the proportion of words recalled\n",
        "- Each condition gets 1 bar\n",
        "- The x-positions of each bar should be in increments of 1, starting with the leftmost bar at 1 (since there are six conditions in the experiment, the rightmost bar should have an x-position of 6).\n",
        "- The bar heights should indicate the average proportion of words recalled in each condition.\n",
        "- Use the `yerr` keyword argument to display confidence intervals (default: `confidence=95`-- i.e., ranging from the bottom 2.5th percentile to the top 2.5th percentile).  The `np.percentile` function may come in handy for this!\n",
        "- Make all bars gray (`color='gray'`) and all lines black (`'k'`)\n",
        "- Set the errorbar capsize to 10 points\n",
        "- Hints:\n",
        "  - You may find it useful to use `pd.DataFrame.apply` to add a new \"proportion recalled\" column to each dataset.  You could do this in a compact way using a lambda function!\n",
        "  - The `spc` function you wrote above can be leveraged to write the `plot_performance_by_condition` function.\n",
        "  - See the [API reference](https://matplotlib.org/3.3.3/api/_as_gen/matplotlib.pyplot.bar.html) for `pyplot.bar` for more information about how to make bar plots using Matplotlib.\n",
        "  - You may find `np.argsort` useful for sorting one list or array based on the elements of another."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oiLn_FxNZHrd"
      },
      "source": [
        "def plot_performance_by_condition(data, confidence=95, outfile=None):\n",
        "  fig, ax = mpl.pyplot.subplots()\n",
        "\n",
        "  ### BEGIN YOUR CODE\n",
        "  pass\n",
        "  ### END YOUR CODE\n",
        "\n",
        "  if outfile is not None:\n",
        "    fig.savefig(outfile)  \n",
        "  return fig, ax"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "mERlzRKoVUxu",
        "outputId": "f44ac8d2-6823-4efa-bc15-61f38da9980b"
      },
      "source": [
        "fig, ax = plot_performance_by_condition(murdock_data);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IL7tpoqSiSdC"
      },
      "source": [
        "# Some closing thoughts\n",
        "\n",
        "Congratulations-- your library tour is complete!  The examples covered in this assignment are really just the tip of the iceberg; each of the libraries you toured may be applied to wide ranges of analyses, datasets, and question spaces.  Becoming fluent in these tools can give you data science superpowers.  Have fun and discover something cool!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b8_1RimggpJI"
      },
      "source": [
        "#used by autograder-- do not modify\n",
        "from PIL import Image\n",
        "\n",
        "def compare_figures(a, b, tolerance=0.9):\n",
        "  '''\n",
        "  Return True iff a and b are the same image, excluding values of 0 and 1, or\n",
        "  transparency (alpha) values.\n",
        "\n",
        "  At least (100*tolorance)% of the values must \"close\" (as measured by\n",
        "  '''\n",
        "  def imread(url, *args, **kwargs):\n",
        "    if os.path.exists(url):\n",
        "      return mpl.pyplot.imread(url)\n",
        "    else:\n",
        "      return np.array(Image.open(get.urlopen(url))) / 255\n",
        "\n",
        "  a = imread(a)[:, :, :3].ravel() #ignore transparency\n",
        "  b = imread(b)[:, :, :3].ravel()\n",
        "\n",
        "  inds = ((a != 0) & (a != 1)) | ((b != 0) & (b != 1))\n",
        "  return (np.sum(np.isclose(a[inds], b[inds])) / np.sum(inds)) >= tolerance"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}