fashion_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6e0d252f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x269c6c95f50>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Datagen inspo taken from \n",
    "# https://colab.research.google.com/github/reiinakano/invariant-risk-minimization/blob/master/invariant_risk_minimization_colored_mnist.ipynb#scrollTo=sopHPgEhu4Jo\n",
    "\n",
    "%matplotlib notebook\n",
    "import os\n",
    "\n",
    "import numpy as np\n",
    "from PIL import Image\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import torch\n",
    "from torch import nn\n",
    "from torchvision import datasets\n",
    "import torchvision.datasets.utils as dataset_utils\n",
    "\n",
    "import torch.optim as optim\n",
    "from torch.utils.data import DataLoader, TensorDataset\n",
    "from torchmetrics import Accuracy, AUROC\n",
    "import time\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import random\n",
    "\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "from sklearn.preprocessing import minmax_scale\n",
    "\n",
    "import pickle \n",
    "import seaborn as sns\n",
    "\n",
    "import scipy\n",
    "from skimage.util import random_noise\n",
    "\n",
    "SEED = 1234\n",
    "DEVICE = 'cpu'\n",
    "np.random.seed(SEED)\n",
    "torch.manual_seed(SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "106d2cbb",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = './data/'\n",
    "if not os.path.exists(data_path):\n",
    "    os.makedirs(data_path)\n",
    "train_mnist = datasets.FashionMNIST(data_path, train=True, download=True)\n",
    "test_mnist = datasets.FashionMNIST(data_path, train=False, download=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1b80bf7c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rescale_values(image,max_val=1.0,min_val=0.0):\n",
    "    '''\n",
    "    image - numpy array\n",
    "    max_val/min_val - float\n",
    "    '''\n",
    "    return (image-image.min())/(image.max()-image.min())*(max_val-min_val)+min_val\n",
    "\n",
    "# Convert MNIST PIL images to 3 channel NP arrays and return separated data and label arrays\n",
    "def preprocess_data(train_mnist, test_mnist):\n",
    "    data = np.zeros((len(train_mnist)+len(test_mnist),28,28,3))\n",
    "    labels = np.zeros(len(train_mnist)+len(test_mnist))\n",
    "    \n",
    "    for i, (img, label) in enumerate(train_mnist):\n",
    "        arr = rescale_values(np.array(img)).reshape((28,28,1))\n",
    "        \n",
    "        im_arr = np.concatenate([np.array(arr),np.zeros((28,28,1)),np.array(arr)], axis=2)\n",
    "    \n",
    "        data[i] = im_arr\n",
    "        labels[i] = label\n",
    "        \n",
    "    for i, (img, label) in enumerate(test_mnist):\n",
    "        arr = rescale_values(np.array(img)).reshape((28,28,1))\n",
    "        \n",
    "        im_arr = np.concatenate([arr,np.zeros((28,28,1)),arr], axis=2)\n",
    "\n",
    "        data[len(train_mnist)+i] = im_arr\n",
    "        labels[len(train_mnist)+i] = label\n",
    "    \n",
    "    return data, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f5fb4bb4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n",
      "0.0\n"
     ]
    }
   ],
   "source": [
    "data, labels = preprocess_data(train_mnist, test_mnist)\n",
    "print(data.max())\n",
    "print(data.min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d59a4112",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this number needs to be picked s.t. 50% is divisible by 10 and 9\n",
    "n_per_class = 3600\n",
    "\n",
    "sorted_data = np.zeros((n_per_class*10,28,28,3))\n",
    "sorted_labels = np.zeros((n_per_class*10,))\n",
    "\n",
    "ind = 0\n",
    "for i in range(10):\n",
    "    class_labels = np.where(labels==i)[0][:n_per_class]\n",
    "    sorted_data[ind:ind+n_per_class] = data[class_labels]\n",
    "    sorted_labels[ind:ind+n_per_class] = labels[class_labels]\n",
    "    \n",
    "    ind += n_per_class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c7c428cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffled 60/20/20 split\n",
    "# split 60/40 into train and val+test, then split val and test by half\n",
    "data_splitter = StratifiedShuffleSplit(n_splits=1, test_size=0.4, random_state=SEED)\n",
    "train_indices, val_indices = list(data_splitter.split(X=sorted_data, y=sorted_labels))[0]\n",
    "\n",
    "x_train = sorted_data[train_indices]\n",
    "y_train = sorted_labels[train_indices]\n",
    "x_val_test = sorted_data[val_indices]\n",
    "y_val_test = sorted_labels[val_indices]\n",
    "\n",
    "data_splitter = StratifiedShuffleSplit(n_splits=1, test_size=0.5, random_state=SEED)\n",
    "val_indices, test_indices = list(data_splitter.split(X=x_val_test, y=y_val_test))[0]\n",
    "\n",
    "x_val = x_val_test[val_indices]\n",
    "y_val = y_val_test[val_indices]\n",
    "\n",
    "x_test = x_val_test[test_indices]\n",
    "y_test = y_val_test[test_indices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a76b327d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1.0, 0, 0.5)\n",
      "(0.9, 0, 0.5)\n",
      "(0.8, 0, 0.5)\n",
      "(0.7, 0, 0.5)\n",
      "(0.6, 0, 0.5)\n",
      "(0.5, 0, 0.5)\n",
      "(0.4, 0, 0.5)\n",
      "(0.3, 0, 0.5)\n",
      "(0.2, 0, 0.5)\n",
      "(0.1, 0, 0.5)\n",
      "(0.0, 0, 0.5)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.5, 0, 0.5)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def create_colours(neutral_scalar=0.5):\n",
    "    red = 1.0\n",
    "    colours = []\n",
    "    for i in range(11):\n",
    "        print((np.round(red,1),0,neutral_scalar))\n",
    "        colours.append((np.round(red,1),0,neutral_scalar))\n",
    "        red -= 0.1\n",
    "        \n",
    "    return colours\n",
    "\n",
    "colours = create_colours()\n",
    "colours.pop(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2628784",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8839dd42",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_device_pixel_ratio', {\n",
       "                device_pixel_ratio: fig.ratio,\n",
       "            });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'dblclick',\n",
       "        on_mouse_event_closure('dblclick')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    fig.rubberband_canvas.style.cursor = msg['cursor'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            var img = evt.data;\n",
       "            if (img.type !== 'image/png') {\n",
       "                /* FIXME: We get \"Resource interpreted as Image but\n",
       "                 * transferred with MIME type text/plain:\" errors on\n",
       "                 * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "                 * to be part of the websocket stream */\n",
       "                img.type = 'image/png';\n",
       "            }\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                img\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from https://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * https://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.key === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.key;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.key !== 'Control') {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    else if (event.altKey && event.key !== 'Alt') {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    else if (event.shiftKey && event.key !== 'Shift') {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k' + event.key;\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pgf\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.binaryType = comm.kernel.ws.binaryType;\n",
       "    ws.readyState = comm.kernel.ws.readyState;\n",
       "    function updateReadyState(_event) {\n",
       "        if (comm.kernel.ws) {\n",
       "            ws.readyState = comm.kernel.ws.readyState;\n",
       "        } else {\n",
       "            ws.readyState = 3; // Closed state.\n",
       "        }\n",
       "    }\n",
       "    comm.kernel.ws.addEventListener('open', updateReadyState);\n",
       "    comm.kernel.ws.addEventListener('close', updateReadyState);\n",
       "    comm.kernel.ws.addEventListener('error', updateReadyState);\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        var data = msg['content']['data'];\n",
       "        if (data['blob'] !== undefined) {\n",
       "            data = {\n",
       "                data: new Blob(msg['buffers'], { type: data['blob'] }),\n",
       "            };\n",
       "        }\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(data);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '149.50392156862745')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# noise = random_noise(, mode='s&p')\n",
    "\n",
    "samp = data[0].copy()\n",
    "samp[:,:,0] *= 0.2\n",
    "samp[:,:,1] *= 0.0\n",
    "samp[:,:,2] *= 0.5\n",
    "\n",
    "\n",
    "fig, axs = plt.subplots(1,4)\n",
    "axs[0].imshow(samp)\n",
    "axs[1].imshow(random_noise(samp[:,:,0], mode='s&p', amount=0.4))\n",
    "axs[1].set_title(np.sum(samp[:,:,0]))\n",
    "\n",
    "axs[2].imshow(samp[:,:,1])\n",
    "axs[2].set_title(np.sum(samp[:,:,1]))\n",
    "\n",
    "axs[3].imshow(random_noise(samp[:,:,2], mode='s&p', amount=0.25))\n",
    "axs[3].set_title(np.sum(samp[:,:,2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "5eac4449",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create suppressor set where 50% of samples of all classes have colour c6=(0.5,0.5,0.5), \n",
    "# and the rest are split evenly 5% for each colour [c0,...,c10]\\{c6}\n",
    "\n",
    "lower = 0\n",
    "upper = 1\n",
    "mu = 0.5\n",
    "sigma = 0.3\n",
    "truncnorm = scipy.stats.truncnorm((lower-mu)/sigma,(upper-mu)/sigma,loc=mu,scale=sigma)\n",
    "\n",
    "def scale_to_sum(arr, arrsum):\n",
    "    return (arrsum / np.sum(np.abs(arr))) * arr\n",
    "\n",
    "def create_suppressor(data, colours, c6=(0.5,0.0,0.5), noise_scale=0.5):\n",
    "    num_samples = len(data)\n",
    "    suppressor_data = np.zeros(data.shape)\n",
    "    colour_labels = np.zeros((data.shape[0],3))\n",
    "    \n",
    "    c6_chunk = data[:int(num_samples/2)]\n",
    "    c6_chunk[:,:,:,0] *= c6[0]\n",
    "    c6_chunk[:,:,:,1] *= c6[1]\n",
    "    c6_chunk[:,:,:,2] *= c6[2]\n",
    "    \n",
    "#     noise = truncnorm.rvs(size=c6_chunk.shape)\n",
    "#     noise[:,:,:,1] *= 0\n",
    "    \n",
    "#     noise[:,:,:,0] = scale_to_sum(noise[:,:,:,0], np.sum(c6_chunk[:,:,:,0]))\n",
    "\n",
    "#     noise[:,:,:,1] = scale_to_sum(noise[:,:,:,1], np.sum(noise[:,:,:,0]*0.5))\n",
    "#     noise[:,:,:,2] = scale_to_sum(noise[:,:,:,2], np.sum(noise[:,:,:,0]*0.5))\n",
    "\n",
    "#     noise[:,:,:,1:] = scale_to_sum(noise[:,:,:,1:], np.sum(noise[:,:,:,0]))\n",
    "    \n",
    "#     c6_chunk = (1-noise_scale)*c6_chunk + abs(noise_scale*noise)\n",
    "    c6_chunk = random_noise(c6_chunk, mode='s&p', amount=noise_scale, seed=SEED)\n",
    "    c6_chunk[:,:,:,1] = 0\n",
    "    \n",
    "    colour_labels[:int(num_samples/2)] = np.array([c6]*int(num_samples/2))\n",
    "    \n",
    "    suppressor_data[:int(num_samples/2)] =  c6_chunk\n",
    "    \n",
    "    for i, data_chunk in enumerate(np.split(data[int(num_samples/2):],10,axis=0)):\n",
    "        chunk_size = len(data_chunk)\n",
    "\n",
    "        data_chunk[:,:,:,0] *= colours[i][0]\n",
    "        data_chunk[:,:,:,1] *= colours[i][1]\n",
    "        data_chunk[:,:,:,2] *= colours[i][2]\n",
    "        \n",
    "#         noise = truncnorm.rvs(size=data_chunk.shape)\n",
    "#         noise[:,:,:,1] *= 0\n",
    "\n",
    "#         data_chunk = (1-noise_scale)*data_chunk + abs(noise_scale*noise)\n",
    "        data_chunk = random_noise(data_chunk, mode='s&p', amount=noise_scale, seed=SEED)\n",
    "        data_chunk[:,:,:,1] = 0\n",
    "        \n",
    "        suppressor_data[int(num_samples/2)+chunk_size*i:int(num_samples/2)+chunk_size*(i+1)] = data_chunk\n",
    "        \n",
    "        colour_labels[int(num_samples/2)+chunk_size*i:int(num_samples/2)+chunk_size*(i+1)] = [colours[i]]*int(data_chunk.shape[0])\n",
    "    return suppressor_data, colour_labels\n",
    "\n",
    "\n",
    "# create confounder set where 50% of samples of all classes have colour c6=(0.5,0.5,0.5), \n",
    "# 25% have the class' colour and the rest are split evenly for each colour [c0,...,c10]\\{c6}\n",
    "def create_confounder(data, class_colour, colours,confounder_portion=0.25, c6=(0.5,0.0,0.5), noise_scale=0.5):\n",
    "    num_samples = len(data)\n",
    "    confounder_data = np.zeros(data.shape)\n",
    "    colour_labels = np.zeros((data.shape[0],3))\n",
    "\n",
    "    c6_chunk = data[:int(num_samples/2)]\n",
    "    c6_chunk[:,:,:,0] *= c6[0]\n",
    "    c6_chunk[:,:,:,1] *= c6[1]\n",
    "    c6_chunk[:,:,:,2] *= c6[2]\n",
    "    \n",
    "#     noise = truncnorm.rvs(size=c6_chunk.shape)\n",
    "#     noise[:,:,:,1] *= 0\n",
    "    \n",
    "#     c6_chunk = (1-noise_scale)*c6_chunk + abs(noise_scale*noise)\n",
    "\n",
    "    c6_chunk = random_noise(c6_chunk, mode='s&p', amount=noise_scale, seed=SEED)\n",
    "    c6_chunk[:,:,:,1] = 0\n",
    "    \n",
    "\n",
    "    colour_labels[:int(num_samples/2)] = np.array([c6]*int(num_samples/2))\n",
    "    confounder_data[:int(num_samples/2)] = c6_chunk\n",
    "    \n",
    "    n_confounded = int(num_samples/2)+int(num_samples*confounder_portion) \n",
    "    \n",
    "    confounder_data[int(num_samples/2):n_confounded] = data[int(num_samples/2):n_confounded]\n",
    "\n",
    "    confounder_data[int(num_samples/2):n_confounded][:,:,:,0] *= class_colour[0]\n",
    "    confounder_data[int(num_samples/2):n_confounded][:,:,:,1] *= class_colour[1]\n",
    "    confounder_data[int(num_samples/2):n_confounded][:,:,:,2] *= class_colour[2]\n",
    "    \n",
    "#     noise = truncnorm.rvs(size=confounder_data[int(num_samples/2):n_confounded].shape)\n",
    "#     noise[:,:,:,1] *= 0\n",
    "    \n",
    "#     confounder_data[int(num_samples/2):n_confounded] = (1-noise_scale)*confounder_data[int(num_samples/2):n_confounded] + abs(noise_scale*noise)\n",
    "    confounder_data[int(num_samples/2):n_confounded] = random_noise(confounder_data[int(num_samples/2):n_confounded], mode='s&p', amount=noise_scale, seed=SEED)\n",
    "    confounder_data[int(num_samples/2):n_confounded][:,:,:,1] = 0\n",
    "    \n",
    "    colour_labels[int(num_samples/2):n_confounded] = np.array([class_colour]*(n_confounded - int(num_samples/2)))\n",
    "    \n",
    "    other_colours = [x for x in colours if x != class_colour] \n",
    "    \n",
    "    for i, data_chunk in enumerate(np.split(data[n_confounded:],9,axis=0)):\n",
    "        chunk_size = len(data_chunk)\n",
    "\n",
    "        data_chunk[:,:,:,0] *= other_colours[i][0]\n",
    "        data_chunk[:,:,:,1] *= other_colours[i][1]\n",
    "        data_chunk[:,:,:,2] *= other_colours[i][2]\n",
    "        \n",
    "#         noise = truncnorm.rvs(size=data_chunk.shape)\n",
    "#         noise[:,:,:,1] *= 0\n",
    "        \n",
    "#         data_chunk = (1-noise_scale)*data_chunk + abs(noise_scale*noise)\n",
    "        data_chunk = random_noise(data_chunk, mode='s&p', amount=noise_scale, seed=SEED)\n",
    "        data_chunk[:,:,:,1] = 0\n",
    "        \n",
    "        confounder_data[n_confounded+chunk_size*i:n_confounded+chunk_size*(i+1)] = data_chunk\n",
    "        \n",
    "        colour_labels[n_confounded+chunk_size*i:n_confounded+chunk_size*(i+1)] = [other_colours[i]]*int(data_chunk.shape[0])\n",
    "        \n",
    "    return confounder_data, colour_labels\n",
    "\n",
    "def create_no_col(data, c6=(0.5,0.0,0.5), noise_scale=0.5):\n",
    "    no_col_data = np.zeros((data.shape))\n",
    "    \n",
    "    c6_chunk = data\n",
    "    c6_chunk[:,:,:,0] *= c6[0]\n",
    "    c6_chunk[:,:,:,1] *= c6[1]\n",
    "    c6_chunk[:,:,:,2] *= c6[2]\n",
    "    \n",
    "#     noise = truncnorm.rvs(size=data.shape)\n",
    "#     noise[:,:,:,1] *= 0\n",
    "    \n",
    "#     c6_chunk = (1-noise_scale)*c6_chunk + abs(noise_scale*noise)\n",
    "    c6_chunk = random_noise(c6_chunk, mode='s&p', amount=noise_scale, seed=SEED)\n",
    "    c6_chunk[:,:,:,1] = 0\n",
    "    \n",
    "    return c6_chunk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "28264be6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# fig, axs = plt.subplots(1,4)\n",
    "\n",
    "# for i in range(4):\n",
    "#     mask = x_train[i].copy()\n",
    "#     mask[mask != 0] = 1\n",
    "#     axs[i].imshow(mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "aef555f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def scale_to_bound(row, scale):\n",
    "    # scale = bound / np.max(np.abs(row))\n",
    "    return row * scale\n",
    "\n",
    "noise_scale = 0.2\n",
    "confounder_portion = 0.4\n",
    "c6_portion = 0.5\n",
    "\n",
    "conf_train = np.zeros(x_train.shape)\n",
    "conf_val = np.zeros(x_val.shape)\n",
    "conf_test = np.zeros(x_test.shape)\n",
    "\n",
    "sup_train = np.zeros(x_train.shape)\n",
    "sup_val = np.zeros(x_val.shape)\n",
    "sup_test = np.zeros(x_test.shape)\n",
    "\n",
    "no_col_train = np.zeros(x_train.shape)\n",
    "no_col_val = np.zeros(x_val.shape)\n",
    "no_col_test = np.zeros(x_test.shape)\n",
    "\n",
    "conf_train_colours = np.zeros((x_train.shape[0], 3))\n",
    "conf_val_colours = np.zeros((x_val.shape[0], 3))\n",
    "conf_test_colours = np.zeros((x_test.shape[0], 3))\n",
    "\n",
    "sup_train_colours = np.zeros((x_train.shape[0], 3))\n",
    "sup_val_colours = np.zeros((x_val.shape[0], 3))\n",
    "sup_test_colours = np.zeros((x_test.shape[0], 3))\n",
    "\n",
    "train_ind = 0\n",
    "val_ind = 0\n",
    "\n",
    "for i in range(10):\n",
    "    train_inds = np.where(y_train==i)[0]\n",
    "    val_inds = np.where(y_val==i)[0]\n",
    "    test_inds = np.where(y_test==i)[0]  \n",
    "\n",
    "    conf_train[train_inds], conf_train_colours[train_inds] = create_confounder(x_train[train_inds].copy(), colours[i], colours, confounder_portion=confounder_portion, noise_scale=noise_scale)\n",
    "    conf_val[val_inds], conf_val_colours[val_inds] = create_confounder(x_val[val_inds].copy(), colours[i], colours, confounder_portion=confounder_portion, noise_scale=noise_scale)\n",
    "    conf_test[test_inds], conf_test_colours[test_inds] = create_confounder(x_test[test_inds].copy(), colours[i], colours, confounder_portion=confounder_portion, noise_scale=noise_scale)\n",
    "    \n",
    "    sup_train[train_inds], sup_train_colours[train_inds] = create_suppressor(x_train[train_inds].copy(), colours, noise_scale=noise_scale)\n",
    "    sup_val[val_inds], sup_val_colours[val_inds] = create_suppressor(x_val[val_inds].copy(), colours, noise_scale=noise_scale)\n",
    "    sup_test[test_inds], sup_test_colours[test_inds] = create_suppressor(x_test[test_inds].copy(), colours, noise_scale=noise_scale)  \n",
    "\n",
    "    no_col_train[train_inds] = create_no_col(x_train[train_inds].copy(), noise_scale=noise_scale)\n",
    "    no_col_val[val_inds] = create_no_col(x_val[val_inds].copy(), noise_scale=noise_scale)\n",
    "    no_col_test[test_inds] = create_no_col(x_test[test_inds].copy(), noise_scale=noise_scale)\n",
    "\n",
    "#     no_col_train[train_inds] = (1.0-noise_scale)*(x_train[train_inds].copy()) + abs(noise_scale*truncnorm.rvs(size=no_col_train[train_inds].shape))\n",
    "#     no_col_val[val_inds] = (1.0-noise_scale)*(x_val[val_inds].copy()) + abs(noise_scale*truncnorm.rvs(size=no_col_val[val_inds].shape))\n",
    "#     no_col_test[test_inds] = (1.0-noise_scale)*(x_test[test_inds].copy()) + abs(noise_scale*truncnorm.rvs(size=no_col_test[test_inds].shape))\n",
    "\n",
    "\n",
    "scale = 1 / np.max(np.abs(np.vstack([conf_train, conf_val, conf_test])))\n",
    "\n",
    "conf_train = np.apply_along_axis(scale_to_bound, 1, conf_train, scale) * 2 - 1\n",
    "conf_val = np.apply_along_axis(scale_to_bound, 1, conf_val, scale) * 2 - 1\n",
    "conf_test = np.apply_along_axis(scale_to_bound, 1, conf_test, scale) * 2 - 1\n",
    "\n",
    "##############\n",
    "\n",
    "scale = 1 / np.max(np.abs(np.vstack([sup_train, sup_val, sup_test])))\n",
    "\n",
    "sup_train = np.apply_along_axis(scale_to_bound, 1, sup_train, scale) * 2 - 1\n",
    "sup_val = np.apply_along_axis(scale_to_bound, 1, sup_val, scale) * 2 - 1\n",
    "sup_test = np.apply_along_axis(scale_to_bound, 1, sup_test, scale) * 2 - 1\n",
    "\n",
    "##############\n",
    "\n",
    "scale = 1 / np.max(np.abs(np.vstack([no_col_train, no_col_val, no_col_test])))\n",
    "\n",
    "no_col_train = np.apply_along_axis(scale_to_bound, 1, no_col_train, scale) * 2 - 1\n",
    "no_col_val = np.apply_along_axis(scale_to_bound, 1, no_col_val, scale) * 2 - 1\n",
    "no_col_test = np.apply_along_axis(scale_to_bound, 1, no_col_test, scale) * 2 - 1\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f3366c50",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-9644732.807132915 -9660874.47716535\n"
     ]
    }
   ],
   "source": [
    "print(np.sum(conf_train[:,:,:,0]), np.sum(conf_train[:,:,:,2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "dacfd9c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-9685142.39649528 -9675729.95924039\n"
     ]
    }
   ],
   "source": [
    "print(np.sum(sup_train[:,:,:,0]), np.sum(sup_train[:,:,:,2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "1557de36",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-9662626.051459007 -9660167.460321154\n"
     ]
    }
   ],
   "source": [
    "print(np.sum(no_col_train[:,:,:,0]), np.sum(no_col_train[:,:,:,2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ad2cdc0f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1.0"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.min(no_col_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b38868bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.max(no_col_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a0ef3380",
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_class_data(x_conf, x_sup, x_no_col, y, class_ind=0, n_cols=10):\n",
    "    class_inds = np.where(y==class_ind)[0]\n",
    "    \n",
    "    sample = random.sample(list(class_inds), n_cols)\n",
    "    fig, axs = plt.subplots(3,n_cols, figsize=(10,4))\n",
    "    \n",
    "    dataset_labels = ['Confounder', 'Suppressor', 'No\\nColour']\n",
    "    \n",
    "    for i, dataset in enumerate([x_conf, x_sup, x_no_col]):\n",
    "        for j in range(n_cols):\n",
    "            if i == 0:\n",
    "                axs[i,j].set_title(f'Sample\\n{sample[j]}')\n",
    "            axs[i,j].imshow(dataset[sample[j]])\n",
    "            axs[i,j].set_xticks([])\n",
    "            axs[i,j].set_yticks([])\n",
    "            \n",
    "        axs[i,0].set_ylabel(dataset_labels[i])\n",
    "            \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "ed80b7c6",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_device_pixel_ratio', {\n",
       "                device_pixel_ratio: fig.ratio,\n",
       "            });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'dblclick',\n",
       "        on_mouse_event_closure('dblclick')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    fig.rubberband_canvas.style.cursor = msg['cursor'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            var img = evt.data;\n",
       "            if (img.type !== 'image/png') {\n",
       "                /* FIXME: We get \"Resource interpreted as Image but\n",
       "                 * transferred with MIME type text/plain:\" errors on\n",
       "                 * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "                 * to be part of the websocket stream */\n",
       "                img.type = 'image/png';\n",
       "            }\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                img\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from https://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * https://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.key === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.key;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.key !== 'Control') {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    else if (event.altKey && event.key !== 'Alt') {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    else if (event.shiftKey && event.key !== 'Shift') {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k' + event.key;\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pgf\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.binaryType = comm.kernel.ws.binaryType;\n",
       "    ws.readyState = comm.kernel.ws.readyState;\n",
       "    function updateReadyState(_event) {\n",
       "        if (comm.kernel.ws) {\n",
       "            ws.readyState = comm.kernel.ws.readyState;\n",
       "        } else {\n",
       "            ws.readyState = 3; // Closed state.\n",
       "        }\n",
       "    }\n",
       "    comm.kernel.ws.addEventListener('open', updateReadyState);\n",
       "    comm.kernel.ws.addEventListener('close', updateReadyState);\n",
       "    comm.kernel.ws.addEventListener('error', updateReadyState);\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        var data = msg['content']['data'];\n",
       "        if (data['blob'] !== undefined) {\n",
       "            data = {\n",
       "                data: new Blob(msg['buffers'], { type: data['blob'] }),\n",
       "            };\n",
       "        }\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(data);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"1000\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    }
   ],
   "source": [
    "show_class_data(conf_train, sup_train, no_col_train, y_train, class_ind=0, n_cols=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "1d069248",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle as pkl\n",
    "\n",
    "conf_save = [(conf_train, y_train, conf_train_colours), (conf_val, y_val, conf_val_colours), (conf_test, y_test, conf_test_colours)]\n",
    "sup_save = [(sup_train, y_train, sup_train_colours), (sup_val, y_val, sup_val_colours), (sup_test, y_test, sup_test_colours)]\n",
    "no_col_save = [(no_col_train, y_train), (no_col_val, y_val), (no_col_test, y_test)]\n",
    "\n",
    "\n",
    "with open('mnist_conf_data.pkl', 'wb') as f:\n",
    "    pkl.dump(conf_save, f)\n",
    "    \n",
    "with open('mnist_sup_data.pkl', 'wb') as f:\n",
    "    pkl.dump(sup_save, f)\n",
    "    \n",
    "with open('mnist_no_col_data.pkl', 'wb') as f:\n",
    "    pkl.dump(no_col_save, f)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6452747f",
   "metadata": {},
   "source": [
    "## Model Training\n",
    "Following the implementation of Anders et. al (2022): https://www.sciencedirect.com/science/article/pii/S1566253521001573\n",
    "and https://colab.research.google.com/github/reiinakano/invariant-risk-minimization/blob/master/invariant_risk_minimization_colored_mnist.ipynb#scrollTo=9hYJRewnv80x\n",
    "\n",
    "\n",
    "All models on colored MNIST in Sections 3.2 and 4.1 are trained\n",
    "using the AdaDelta algorithm with a learning rate of 1.0, which is\n",
    "multiplied by 0.7 after each epoch, for 10 epochs. The a posteriori ClArC\n",
    "is trained for 10 epochs on top of the baseline model, which has also\n",
    "been trained for 10 epochs. The network consists of 2 convolutional\n",
    "layers, followed by a max-pooling, and finally 2 fully-connected layers.\n",
    "Dropout is used after the max pooling and after the first fully-connected\n",
    "layer, with 25 percent and 50 percent dropout probabilities respectively. ReLU activations follow all fully-connected layers except the\n",
    "final one. The model used for 5.1 is trained with SGD, with a learning\n",
    "rate of 0.001 for 5 epochs. The architecture, however, is the same as\n",
    "for the other colored MNIST models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "b5631e99",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Net(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Net, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(3, 20, 5, 1)\n",
    "        self.conv2 = nn.Conv2d(20, 50, 5, 1)\n",
    "        self.fc1 = nn.Linear(4 * 4 * 50, 500)\n",
    "        self.fc2 = nn.Linear(500, 10)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.conv1(x))\n",
    "        x = F.max_pool2d(x, 2, 2)\n",
    "        x = F.relu(self.conv2(x))\n",
    "        x = F.max_pool2d(x, 2, 2)\n",
    "        x = x.view(-1, 4 * 4 * 50)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        logits = self.fc2(x)\n",
    "        return logits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "a3a5d77d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(lr,epochs,train_loader,val_loader,name_model,momentum=0.9,weight_decay=1e-3,num_models=10):\n",
    "    metric = AUROC(task='multiclass', num_classes=10) \n",
    "    softmax = nn.Softmax(dim=1)\n",
    "    \n",
    "    training_accuracy = Accuracy(task='multiclass', num_classes=10).to(DEVICE)\n",
    "    val_accuracy = Accuracy(task='multiclass', num_classes=10).to(DEVICE)\n",
    "\n",
    "    torch.manual_seed(SEED)\n",
    "\n",
    "    trained_models = []\n",
    "    training_logs = []\n",
    "\n",
    "    for training_ind in range(num_models):\n",
    "        cnn_conf=Net()\n",
    "        cnn_conf.to(DEVICE)\n",
    "        criterion = nn.CrossEntropyLoss()\n",
    "        \n",
    "        # Anders et. al uses mainly AdaDelta but also SGD in one section\n",
    "#         optimizer = optim.SGD(cnn_conf.parameters(), lr=lr, momentum=momentum, weight_decay=weight_decay)\n",
    "        optimizer = optim.Adadelta(cnn_conf.parameters(), lr=lr, weight_decay=weight_decay)\n",
    "\n",
    "\n",
    "        loss_list_train=[]\n",
    "        loss_list_val=[]\n",
    "        acc_list_train=[]\n",
    "        acc_list_val=[]\n",
    "        \n",
    "\n",
    "        min_loss = 100000000\n",
    "    #     last_loss = 1000000000\n",
    "\n",
    "        for epoch in range(epochs):  # loop over the dataset multiple times\n",
    "            print(f' epoch {epoch+1} in {epochs}')\n",
    "            t0=time.time()\n",
    "            epoch_loss = 0.0\n",
    "            epoch_loss_val= 0.0\n",
    "            batch_acc=[]\n",
    "            batch_acc_val=[]\n",
    "\n",
    "            auroc_train=[]\n",
    "            auroc_val=[]\n",
    "\n",
    "            pred_train=[]\n",
    "            pred_val=[]\n",
    "            true_labels_train=[]\n",
    "            true_labels_val=[]\n",
    "\n",
    "            for i, data in enumerate(train_loader):\n",
    "                inputs, labels = data\n",
    "\n",
    "                inputs = inputs.to(DEVICE,dtype=torch.float)\n",
    "                labels = labels.type(torch.LongTensor)\n",
    "                labels=labels.to(DEVICE)\n",
    "\n",
    "                # zero the parameter gradients\n",
    "                optimizer.zero_grad()\n",
    "\n",
    "                # forward + backward + optimize\n",
    "                outputs = cnn_conf(inputs).squeeze()\n",
    "                _, predicted = torch.max(outputs.data,1)\n",
    "                loss = criterion(outputs, labels)\n",
    "                loss.backward()\n",
    "                optimizer.step()\n",
    "\n",
    "                epoch_loss += loss.item()        \n",
    "                batch_acc.append(training_accuracy(predicted, labels).item())\n",
    "\n",
    "#                 print(softmax(outputs)[:,1].detach().numpy().shape)\n",
    "#                 pred_train=np.concatenate((pred_train, softmax(outputs)[:,1].detach().numpy()))\n",
    "#                 true_labels_train=np.concatenate((true_labels_train, labels.cpu().numpy()))\n",
    "                \n",
    "                auroc_train.append(metric(outputs,labels.to(torch.int32)))\n",
    "\n",
    "            auroc_train = sum(auroc_train)/len(auroc_train)\n",
    "\n",
    "            for i_v, data_val in enumerate(val_loader):\n",
    "                inputs_val, labels_val = data_val\n",
    "\n",
    "                inputs_val = inputs_val.to(DEVICE,dtype=torch.float)\n",
    "                labels_val = labels_val.type(torch.LongTensor)\n",
    "                labels_val=labels_val.to(DEVICE)\n",
    "\n",
    "                outputs_val = cnn_conf(inputs_val).squeeze()\n",
    "                _, predicted_val = torch.max(outputs_val.data, 1)\n",
    "\n",
    "                loss_val = criterion(outputs_val, labels_val)\n",
    "\n",
    "                epoch_loss_val += loss_val.item()        \n",
    "                batch_acc_val.append(val_accuracy(predicted_val, labels_val).item())\n",
    "\n",
    "                auroc_val.append(metric(outputs_val,labels_val.to(torch.int32)))\n",
    "\n",
    "#                 pred_val=np.concatenate((pred_val, softmax(outputs_val)[:,1].cpu().detach().numpy()))\n",
    "#                 true_labels_val=np.concatenate((true_labels_val, labels_val.cpu().detach().numpy()))\n",
    "            auroc_val = sum(auroc_val)/len(auroc_val)\n",
    "\n",
    "            if epoch_loss_val < min_loss:\n",
    "                min_loss = epoch_loss_val\n",
    "                save_model = type(cnn_conf)() # get a new instance\n",
    "                save_model.load_state_dict(cnn_conf.state_dict()) # copy weights and stuff\n",
    "\n",
    "                torch.save(save_model.state_dict(), f'{str(name_model)}_{training_ind}.pt')\n",
    "\n",
    "#             auroc_val=metric(torch.tensor(pred_val),torch.tensor(true_labels_val).to(torch.int32))\n",
    "\n",
    "            epoch_acc=sum(batch_acc)/len(batch_acc)\n",
    "            epoch_acc_val=sum(batch_acc_val)/len(batch_acc_val)\n",
    "\n",
    "            print(f'epoch train loss: {epoch_loss} | epoch train acc {epoch_acc} | AUROC: {auroc_train}')\n",
    "            print(f'epoch val loss: {epoch_loss_val} | epoch val acc {epoch_acc_val} | AUROC: {auroc_val}')\n",
    "            print(f'time elapsed: {round(time.time()-t0,2)} s')\n",
    "\n",
    "            loss_list_train.append(epoch_loss)\n",
    "            loss_list_val.append(epoch_loss_val)\n",
    "            acc_list_train.append(epoch_acc)\n",
    "            acc_list_val.append(epoch_acc_val)\n",
    "\n",
    "            epoch_loss = 0.0\n",
    "            epoch_loss_val=0.0\n",
    "\n",
    "        print('Finished Training')\n",
    "        print()\n",
    "        print('\\t \\t *******************')\n",
    "        print()\n",
    "        \n",
    "        torch.save(save_model.state_dict(), f'{str(name_model)}_{training_ind}.pt')\n",
    "\n",
    "        trained_models.append(save_model)\n",
    "        training_logs.append([acc_list_train, loss_list_train, acc_list_val, loss_list_val])\n",
    "#     plt.plot(loss_list_train)\n",
    "#     plt.plot(loss_list_val)\n",
    "#     plt.plot([i*100 for i in acc_list_train])\n",
    "#     plt.plot([i*100 for i in acc_list_va])\n",
    "    return trained_models, [acc_list_train, loss_list_train, acc_list_val, loss_list_val]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "9415ca55",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader = DataLoader(TensorDataset(torch.tensor(conf_train.transpose(0,3,1,2)), torch.tensor(y_train)), batch_size=1000, shuffle=True)\n",
    "val_loader = DataLoader(TensorDataset(torch.tensor(conf_val.transpose(0,3,1,2)), torch.tensor(y_val)), batch_size=1000, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "4d40a193",
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = 0.005\n",
    "epochs = 50\n",
    "num_models=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "1857af32",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " epoch 1 in 50\n",
      "epoch train loss: 50.67912459373474 | epoch train acc 0.10334848409349268 | AUROC: 0.5045902729034424\n",
      "epoch val loss: 18.413030862808228 | epoch val acc 0.10737500060349703 | AUROC: 0.5468242168426514\n",
      "time elapsed: 8.09 s\n",
      " epoch 2 in 50\n",
      "epoch train loss: 50.60353183746338 | epoch train acc 0.11583333293145354 | AUROC: 0.5760854482650757\n",
      "epoch val loss: 18.38860297203064 | epoch val acc 0.12325000111013651 | AUROC: 0.6111649870872498\n",
      "time elapsed: 10.09 s\n",
      " epoch 3 in 50\n",
      "epoch train loss: 50.52973222732544 | epoch train acc 0.14539393951947038 | AUROC: 0.6387097239494324\n",
      "epoch val loss: 18.36316227912903 | epoch val acc 0.15512499772012234 | AUROC: 0.6662144660949707\n",
      "time elapsed: 10.94 s\n",
      " epoch 4 in 50\n",
      "epoch train loss: 50.457945585250854 | epoch train acc 0.16677272658456455 | AUROC: 0.6867557764053345\n",
      "epoch val loss: 18.333980321884155 | epoch val acc 0.17762500047683716 | AUROC: 0.7112770676612854\n",
      "time elapsed: 13.34 s\n",
      " epoch 5 in 50\n",
      "epoch train loss: 50.38235950469971 | epoch train acc 0.18392424420876938 | AUROC: 0.724199116230011\n",
      "epoch val loss: 18.309396505355835 | epoch val acc 0.18599999882280827 | AUROC: 0.7399702072143555\n",
      "time elapsed: 11.08 s\n",
      " epoch 6 in 50\n",
      "epoch train loss: 50.30474591255188 | epoch train acc 0.18768181651830673 | AUROC: 0.7508319616317749\n",
      "epoch val loss: 18.274803638458252 | epoch val acc 0.18662500195205212 | AUROC: 0.7628026008605957\n",
      "time elapsed: 11.54 s\n",
      " epoch 7 in 50\n",
      "epoch train loss: 50.216694831848145 | epoch train acc 0.18836363675919446 | AUROC: 0.7711272835731506\n",
      "epoch val loss: 18.24385404586792 | epoch val acc 0.1823749989271164 | AUROC: 0.7765519618988037\n",
      "time elapsed: 11.75 s\n",
      " epoch 8 in 50\n",
      "epoch train loss: 50.12508273124695 | epoch train acc 0.18578788096254523 | AUROC: 0.7862890362739563\n",
      "epoch val loss: 18.210386514663696 | epoch val acc 0.1808749996125698 | AUROC: 0.7920438647270203\n",
      "time elapsed: 12.13 s\n",
      " epoch 9 in 50\n",
      "epoch train loss: 50.01873850822449 | epoch train acc 0.18495454503731293 | AUROC: 0.7988593578338623\n",
      "epoch val loss: 18.166194438934326 | epoch val acc 0.1848750002682209 | AUROC: 0.8007956743240356\n",
      "time elapsed: 11.29 s\n",
      " epoch 10 in 50\n",
      "epoch train loss: 49.89622449874878 | epoch train acc 0.18542423979802566 | AUROC: 0.8085504174232483\n",
      "epoch val loss: 18.11971426010132 | epoch val acc 0.18412500247359276 | AUROC: 0.8122260570526123\n",
      "time elapsed: 12.17 s\n",
      " epoch 11 in 50\n",
      "epoch train loss: 49.7615647315979 | epoch train acc 0.18434848500923676 | AUROC: 0.8165708780288696\n",
      "epoch val loss: 18.05177593231201 | epoch val acc 0.19099999964237213 | AUROC: 0.8212369680404663\n",
      "time elapsed: 10.67 s\n",
      " epoch 12 in 50\n",
      "epoch train loss: 49.60044527053833 | epoch train acc 0.18401515212925998 | AUROC: 0.8233718872070312\n",
      "epoch val loss: 18.001407623291016 | epoch val acc 0.1815000008791685 | AUROC: 0.8273824453353882\n",
      "time elapsed: 10.48 s\n",
      " epoch 13 in 50\n",
      "epoch train loss: 49.4092857837677 | epoch train acc 0.18634848499839957 | AUROC: 0.8292011022567749\n",
      "epoch val loss: 17.92261052131653 | epoch val acc 0.190250001847744 | AUROC: 0.8304693102836609\n",
      "time elapsed: 10.96 s\n",
      " epoch 14 in 50\n",
      "epoch train loss: 49.18802499771118 | epoch train acc 0.19713636419989847 | AUROC: 0.8336509466171265\n",
      "epoch val loss: 17.82796621322632 | epoch val acc 0.19900000095367432 | AUROC: 0.8339409828186035\n",
      "time elapsed: 13.51 s\n",
      " epoch 15 in 50\n",
      "epoch train loss: 48.92153334617615 | epoch train acc 0.20337878709489648 | AUROC: 0.8371169567108154\n",
      "epoch val loss: 17.71930432319641 | epoch val acc 0.20662500150501728 | AUROC: 0.8391349911689758\n",
      "time elapsed: 10.59 s\n",
      " epoch 16 in 50\n",
      "epoch train loss: 48.60294151306152 | epoch train acc 0.22178787873549896 | AUROC: 0.8407294750213623\n",
      "epoch val loss: 17.594515085220337 | epoch val acc 0.21975000016391277 | AUROC: 0.8435681462287903\n",
      "time elapsed: 15.48 s\n",
      " epoch 17 in 50\n",
      "epoch train loss: 48.22145676612854 | epoch train acc 0.22766666656190698 | AUROC: 0.844004213809967\n",
      "epoch val loss: 17.445221424102783 | epoch val acc 0.2322499994188547 | AUROC: 0.8435453772544861\n",
      "time elapsed: 13.6 s\n",
      " epoch 18 in 50\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\1\\ipykernel_24516\\1339579275.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mtrain\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlr\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mepochs\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtrain_loader\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mval_loader\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;34m'mnist_conf'\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mmomentum\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0.9\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mweight_decay\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m1e-3\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mnum_models\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnum_models\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\1\\ipykernel_24516\\377032434.py\u001b[0m in \u001b[0;36mtrain\u001b[1;34m(lr, epochs, train_loader, val_loader, name_model, momentum, weight_decay, num_models)\u001b[0m\n\u001b[0;32m     60\u001b[0m                 \u001b[0m_\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpredicted\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     61\u001b[0m                 \u001b[0mloss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlabels\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 62\u001b[1;33m                 \u001b[0mloss\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     63\u001b[0m                 \u001b[0moptimizer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     64\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\.virtualenvs\\debugging_xai-9luXJGIR\\lib\\site-packages\\torch\\_tensor.py\u001b[0m in \u001b[0;36mbackward\u001b[1;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[0;32m    487\u001b[0m             )\n\u001b[0;32m    488\u001b[0m         torch.autograd.backward(\n\u001b[1;32m--> 489\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mgradient\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0minputs\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    490\u001b[0m         )\n\u001b[0;32m    491\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\.virtualenvs\\debugging_xai-9luXJGIR\\lib\\site-packages\\torch\\autograd\\__init__.py\u001b[0m in \u001b[0;36mbackward\u001b[1;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[0;32m    197\u001b[0m     Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass\n\u001b[0;32m    198\u001b[0m         \u001b[0mtensors\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mgrad_tensors_\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 199\u001b[1;33m         allow_unreachable=True, accumulate_grad=True)  # Calls into the C++ engine to run the backward pass\n\u001b[0m\u001b[0;32m    200\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    201\u001b[0m def grad(\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "train(lr,epochs,train_loader,val_loader,'mnist_conf',momentum=0.9,weight_decay=1e-3,num_models=num_models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "59fffd1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader = DataLoader(TensorDataset(torch.tensor(sup_train.transpose(0,3,1,2)), torch.tensor(y_train)), batch_size=1000, shuffle=True)\n",
    "val_loader = DataLoader(TensorDataset(torch.tensor(sup_val.transpose(0,3,1,2)), torch.tensor(y_val)), batch_size=1000, shuffle=True)\n",
    "\n",
    "train(lr,epochs,train_loader,val_loader,'mnist_sup',momentum=0.9,weight_decay=1e-3,num_models=num_models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c7bf2bac",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader = DataLoader(TensorDataset(torch.tensor(no_col_train.transpose(0,3,1,2)), torch.tensor(y_train)), batch_size=1000, shuffle=True)\n",
    "val_loader = DataLoader(TensorDataset(torch.tensor(no_col_val.transpose(0,3,1,2)), torch.tensor(y_val)), batch_size=1000, shuffle=True)\n",
    "\n",
    "train(lr,epochs,train_loader,val_loader,'mnist_no_col',momentum=0.9,weight_decay=1e-3,num_models=num_models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "12cef5c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_trained(path, device='cpu'):\n",
    "    model = Net()\n",
    "    model.load_state_dict(torch.load(path,map_location=device))\n",
    "    return model\n",
    "\n",
    "def print_AUC(loader, model_ind=0, device='cpu'):\n",
    "\n",
    "    cnn_conf=load_trained(f'./mnist_conf_{model_ind}.pt').to(device)\n",
    "    cnn_sup=load_trained(f'./mnist_sup_{model_ind}.pt').to(device)\n",
    "    cnn_no=load_trained(f'./mnist_no_col_{model_ind}.pt').to(device)\n",
    "    \n",
    "    #models -> cnn_no; cnn_sup; cnn_conf\n",
    "\n",
    "    softmax = nn.Softmax(dim=1)\n",
    "    metric = AUROC(num_classes=10, task='multiclass')\n",
    "\n",
    "    model_conf_test_conf_pred=np.array([])\n",
    "    model_sup_test_conf_pred=np.array([])\n",
    "    model_no_test_conf_pred=np.array([])\n",
    "    true_labels=np.array([])\n",
    "    \n",
    "    model_conf_auroc_test = []\n",
    "    model_sup_auroc_test = []\n",
    "    model_no_col_auroc_test = []\n",
    "\n",
    "    for i_v, data_val in enumerate(loader):\n",
    "        inputs_val, labels_val = data_val\n",
    "        inputs_val = inputs_val.to(DEVICE,dtype=torch.float)\n",
    "        labels_val = labels_val.type(torch.LongTensor)\n",
    "        labels_val=labels_val.to(DEVICE)\n",
    "\n",
    "        outputs_conf = cnn_conf(inputs_val).squeeze()\n",
    "        outputs_sup = cnn_sup(inputs_val).squeeze()\n",
    "        outputs_no = cnn_no(inputs_val).squeeze()\n",
    "\n",
    "        out_pred_conf=softmax(outputs_conf)\n",
    "        out_pred_sup=softmax(outputs_sup)\n",
    "        out_pred_no=softmax(outputs_no)\n",
    "\n",
    "#         model_conf_test_conf_pred=np.concatenate((model_conf_test_conf_pred, out_pred_conf[:,1].cpu().detach().numpy()))\n",
    "#         model_sup_test_conf_pred=np.concatenate((model_sup_test_conf_pred, out_pred_sup[:,1].cpu().detach().numpy()))\n",
    "#         model_no_test_conf_pred=np.concatenate((model_no_test_conf_pred, out_pred_no[:,1].cpu().detach().numpy()))\n",
    "\n",
    "#         print(labels_val.shape)\n",
    "#         print(model_conf_test_conf_pred)\n",
    "\n",
    "        model_conf_auroc_test.append(metric(out_pred_conf, labels_val))\n",
    "        model_sup_auroc_test.append(metric(out_pred_sup, labels_val))\n",
    "        model_no_col_auroc_test.append(metric(out_pred_no, labels_val))\n",
    "        \n",
    "#         true_labels=np.concatenate((true_labels, labels_val.cpu().numpy()))\n",
    "\n",
    "    print('model: confounder', sum(model_conf_auroc_test)/len(model_conf_auroc_test))\n",
    "    print('model: suppressor',sum(model_sup_auroc_test)/len(model_sup_auroc_test))\n",
    "    print('model: no colour',sum(model_no_col_auroc_test)/len(model_no_col_auroc_test))\n",
    "    print()\n",
    "\n",
    "    l=[sum(model_conf_auroc_test)/len(model_conf_auroc_test),\n",
    "       sum(model_sup_auroc_test)/len(model_sup_auroc_test),\n",
    "       sum(model_no_col_auroc_test)/len(model_no_col_auroc_test)]\n",
    "    return l\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "baad3499",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle as pkl\n",
    "with open('mnist_conf_data.pkl', 'rb') as f:\n",
    "    conf_data = pkl.load(f)\n",
    "    \n",
    "with open('mnist_sup_data.pkl', 'rb') as f:\n",
    "    sup_data = pkl.load(f)\n",
    "    \n",
    "with open('mnist_no_col_data.pkl', 'rb') as f:\n",
    "    no_col_data = pkl.load(f)\n",
    "\n",
    "batch_size = 64\n",
    "\n",
    "confounder_test_loader = DataLoader(TensorDataset(torch.tensor(conf_data[2][0].transpose(0,3,1,2)), torch.tensor(conf_data[2][1])),batch_size=batch_size, shuffle=True)\n",
    "suppressor_test_loader = DataLoader(TensorDataset(torch.tensor(sup_data[2][0].transpose(0,3,1,2)), torch.tensor(sup_data[2][1])),batch_size=batch_size, shuffle=True)\n",
    "no_col_test_loader = DataLoader(TensorDataset(torch.tensor(no_col_data[2][0].transpose(0,3,1,2)), torch.tensor(no_col_data[2][1])),batch_size=batch_size, shuffle=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "9c4b2f38",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MODEL 0\n",
      "confounder data:\n",
      "model: confounder tensor(0.9720)\n",
      "model: suppressor tensor(0.9687)\n",
      "model: no colour tensor(0.9689)\n",
      "\n",
      "\n",
      "suppressor data:\n",
      "model: confounder tensor(0.9624)\n",
      "model: suppressor tensor(0.9683)\n",
      "model: no colour tensor(0.9696)\n",
      "\n",
      "\n",
      "no colour data:\n",
      "model: confounder tensor(0.9700)\n",
      "model: suppressor tensor(0.9722)\n",
      "model: no colour tensor(0.9745)\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "batch_size=64\n",
    "# confounder_test_loader = DataLoader(TensorDataset(torch.tensor(conf_test.transpose(0,3,1,2)), torch.tensor(y_test)),batch_size=batch_size, shuffle=False)\n",
    "# supressor_test_loader = DataLoader(TensorDataset(torch.tensor(sup_test.transpose(0,3,1,2)), torch.tensor(y_test)),batch_size=batch_size, shuffle=False)\n",
    "# no_col_test_loader = DataLoader(TensorDataset(torch.tensor(no_col_test.transpose(0,3,1,2)), torch.tensor(y_test)),batch_size=batch_size, shuffle=False)\n",
    "\n",
    "confounder_data_results = []\n",
    "suppressor_data_results = []\n",
    "no_col_data_results = []\n",
    "\n",
    "for i in range(1):\n",
    "    print(f'MODEL {i}')\n",
    "    print('confounder data:')\n",
    "    confounder_data_results.append(print_AUC(confounder_test_loader, i))\n",
    "    print()\n",
    "    print('suppressor data:')\n",
    "    suppressor_data_results.append(print_AUC(suppressor_test_loader, i))\n",
    "    print()\n",
    "    print('no colour data:')\n",
    "    no_col_data_results.append(print_AUC(no_col_test_loader, i))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3a135d8d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# current problem\n",
    "# scaling channel noise by sum of that channel will colour background in accordance with pattern colour\n",
    "# alternative: scale channel noise according to another factor, i.e. sum(channel_0) = 2*sum(channel_1 or _2)\n",
    "# key aim: keep total mass in red = green + blue\n",
    "# another problem: we have the c6_chunk = (1.0, 1.0, 1.0) where r = g + b; should we just make this 1.0, 0.5, 0.5? it's "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "debugging-xai",
   "language": "python",
   "name": "debugging-xai"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}