tutorials/explainers/RISE/rise_timeseries_frb.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "<img width=\"150\" alt=\"Logo_ER10\" src=\"https://user-images.githubusercontent.com/3244249/151994514-b584b984-a148-4ade-80ee-0f88b0aefa45.png\">\n",
    "\n",
    "### Model interpretation using RISE for astronomical timeseries data\n",
    "This notebook shows how to apply the RISE explainability method on a model trained to classify Fast Radio Burst (FRB) timeseries data. The model is a binary classifier that classifies the input as either noise or a real FRB.<br>\n",
    "\n",
    "[RISE](http://bmvc2018.org/contents/papers/1064.pdf) is short for Randomized Input Sampling for Explanation of Black-box Models. It estimates the relevance empirically by probing the model with randomly masked versions of the input image to obtain the corresponding outputs.<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Colab Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-30T16:49:27.348309Z",
     "start_time": "2023-05-30T16:49:16.050787Z"
    }
   },
   "outputs": [],
   "source": [
    "running_in_colab = 'google.colab' in str(get_ipython())\n",
    "if running_in_colab:\n",
    "    # install dianna\n",
    "    !python3 -m pip install dianna[notebooks]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 0 - Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-05-30T16:49:27.399443Z",
     "start_time": "2023-05-30T16:49:17.537497Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "import dianna\n",
    "from dianna import visualization\n",
    "from dianna.utils import SimpleModelRunner\n",
    "from dianna.utils.downloader import download\n",
    "\n",
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_path = download('apertif_frb_dynamic_spectrum_model.onnx', 'model')\n",
    "classes_path = download('apertif_frb_classes.txt', 'label')\n",
    "data_path = download('FRB211024.npy', 'data')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1- Load and visualize the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = np.load(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'FRB 211024')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ax.imshow(data, aspect='auto', origin='lower')\n",
    "ax.set_xlabel('Time step')\n",
    "ax.set_ylabel('Channel index')\n",
    "ax.set_title('FRB 211024')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the figure, we see mostly noise, with a bright vertical line in the center. The vertical line is the fast radio burst (FRB) signal. Different channels correspond to different frequencies of light. Some channels are known to contain bad data and are set to zero."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The classification task\n",
    "\n",
    "The goal for a neural network is to determine whether or not the input data contains an FRB-like signal. Hence, it is a binary classification task. The two classes are noise and FRB. The model used here is a convolutional neural network, described further in [this paper](https://doi.org/10.3847/1538-3881/aae649). The model is automatically downloaded by this notebook, and available [here](https://zenodo.org/record/10656614).\n",
    "\n",
    "Note that the data are timeseries, while a CNN is typically used for images. It works here because neighboring steps in both the time and channel axes are correlated, as would be the case for an image."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Select the class to run the explainer for"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class_names=['Noise', 'FRB']\n",
      "Will generate explanation for class FRB\n"
     ]
    }
   ],
   "source": [
    "with open(classes_path) as fp:\n",
    "    class_names = [line.strip() for line in fp.readlines()]\n",
    "    \n",
    "print(f'{class_names=}')\n",
    "class_to_explain = 'FRB'\n",
    "labels = [class_names.index(class_to_explain)]\n",
    "print(f'Will generate explanation for class {class_to_explain}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define function to handle difference between data shape expected by DIANNA and model\n",
    "\n",
    "Dianna expects input with shape `(time, channel)` while the model and data use `(channel, time)`\n",
    "In addition, the model needs an extra emtpy axis at the end. \n",
    "(model detail: this is because the model is a CNN and the input is processed as if it is an image. The empty axis at the end is the \"colour channel\" axis)\n",
    "\n",
    "To handle this, we transpose the input data before feeding it to DIANNA and we define a preprocess function that is called on the data just before it is given to the model. DIANNA creates masked data in batches, so the preprocess function receives data with shape `(batch, time, channel)` which we need to convert to `(batch, channel, time, 1)`. We also convert to `float32`, as this is the only floating point format handled by models in ONNX format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess(data):\n",
    "    return np.transpose(data, (0, 2, 1))[..., None].astype(np.float32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2 - Run the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Noise: 2.78e-29    FRB: 1.00e+00    \n"
     ]
    }
   ],
   "source": [
    "def run_model(input_data):\n",
    "    \"\"\"Run the model for a single data item (no batch axis)\n",
    "    \"\"\"\n",
    "    runner = SimpleModelRunner(model_path)\n",
    "    model_output = runner(preprocess(input_data.T[None, ...]))[0]\n",
    "    for idx, value in enumerate(model_output):\n",
    "        print(f'{class_names[idx]}: {value:.2e}', end='    ')\n",
    "    print()\n",
    "\n",
    "run_model(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the model is quite sure that the input data contains an FRB. This is indeed correct, however we want to know if the model indeed uses the actual FRB signal (the vertical line in the center of the input data) to come to this conclusion. We will now run the DIANNA timeseries explainer to explore this"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "#### 3 - Run the explainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Explaining: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 50/50 [00:01<00:00, 38.79it/s]\n"
     ]
    }
   ],
   "source": [
    "def run_explainer(input_data):\n",
    "    \"\"\"Run DIANNA timeseries explainer with fixed parameters\n",
    "    Returns a heatmap of the same shape as the input data\n",
    "    \"\"\"\n",
    "    # Run the explainer with 5000 masks and a feature resolution of 8 samples, with a 50% chance of keeping a sample in each mask\n",
    "    n_masks = 5000\n",
    "    feature_res = input_data.shape[1] // 4  # feature size is roughly 4 timesteps\n",
    "    p_keep = .1\n",
    "\n",
    "    # Masked values are replaced by the mean of the input data\n",
    "    explanation = dianna.explain_timeseries(model_path, input_data.T, method='RISE', labels=labels,\n",
    "                                            n_masks=n_masks, feature_res=feature_res, p_keep=p_keep,\n",
    "                                            preprocess_function=preprocess, mask_type='mean')\n",
    "\n",
    "\n",
    "    # select heatmap of the first (and only) class we ran the explainer for\n",
    "    heatmap = explanation[0].T\n",
    "    return heatmap\n",
    "\n",
    "heatmap = run_explainer(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4 -  Plot the explanation and input data next to each other"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_data_and_heatmap(input_data, output_heatmap):\n",
    "    fig, axes = plt.subplots(ncols=2, figsize=(14, 6))\n",
    "\n",
    "    ax = axes[0]\n",
    "    ax.imshow(input_data, aspect='auto', origin='lower')\n",
    "    ax.set_xlabel('Time step')\n",
    "    ax.set_ylabel('Channel index')\n",
    "    ax.set_title('Input data')\n",
    "\n",
    "    ax = axes[1]\n",
    "    ax.imshow(output_heatmap, aspect='auto', origin='lower', cmap='bwr')\n",
    "    ax.set_xlabel('Time step')\n",
    "    ax.set_ylabel('Channel index')\n",
    "    ax.set_title(f'Explanation for class {class_to_explain}')\n",
    "    \n",
    "plot_data_and_heatmap(data, heatmap)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 5 -  Results\n",
    "\n",
    "In the explainer output, red means in favour of the given class and blue means against. From the output, we can clearly see that the model is classifying this as an FRB because of the vertical line in the center, which is exactly where the real FRB is located so in this case the model behaves as expected.\n",
    "\n",
    "#### 6 - Modify the input data\n",
    "A bright FRB right in the center of the input data is perhaps the easiest for the model to classify correctly, as this is exactly what it was trained for. We now explore two ways the input data may be affected:\n",
    "\n",
    "1. We can move the FRB signal away from the center of the input data.\n",
    "2. To simulate deleterious signals from other radio sources on Earth, we can add a bright signal to a few channels that is always on.\n",
    "\n",
    "\n",
    "#### Move the FRB signal away from the center"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Noise: 2.81e-08    FRB: 1.00e+00    \n"
     ]
    }
   ],
   "source": [
    "# Shift the FRB 10 timesteps to the right\n",
    "data_shifted = np.roll(data, 10)\n",
    "run_model(data_shifted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "heatmap_shifted = run_explainer(data_shifted)\n",
    "plot_data_and_heatmap(data_shifted, heatmap_shifted)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the model handles a modest shift fine, and still uses the real location of the FRB in the data in the classification. However, the model fails once the FRB is too far from the center as we can see in the following example. This may be beacuse in the training data, the FRB was never this far from the center."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Noise: 1.00e+00    FRB: 7.67e-18    \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Explaining: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 50/50 [00:01<00:00, 43.28it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Shift the FRB 15 timesteps to the right\n",
    "data_shifted_too_far = np.roll(data, 15)\n",
    "run_model(data_shifted_too_far)\n",
    "\n",
    "heatmap_shifted_too_far = run_explainer(data_shifted_too_far)\n",
    "plot_data_and_heatmap(data_shifted_too_far, heatmap_shifted_too_far)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model now wrongly classifies the input as noise, and we see the the FRB signal is acutally the main reason why the model is _against_ this being an FRB."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Add a deleterious signal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Noise: 4.37e-13    FRB: 1.00e+00    \n"
     ]
    }
   ],
   "source": [
    "data_with_signal = data.copy()\n",
    "\n",
    "# we add a deleterious signal that is slightly stronger than the real FRB\n",
    "rand_mean = 1.5 * np.amax(data)\n",
    "rand_std = data.std()\n",
    "data_with_signal[15] = np.random.normal(rand_mean, rand_std, size=data_with_signal.shape[1])\n",
    "run_model(data_with_signal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Explaining: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 50/50 [00:01<00:00, 37.05it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "heatmap_with_signal = run_explainer(data_with_signal)\n",
    "plot_data_and_heatmap(data_with_signal, heatmap_with_signal)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the output we can see both the real FRB (vertical line) and deleterious signal (horizontal line). In this case, the model still recognizes the FRB. However, we can also see that the deleterious signal does push the model away from the class FRB.\n",
    "\n",
    "As for the previous example, we can check at which point the model breaks down. Let's rerun with more channels affected by a bad signal."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Noise: 1.00e+00    FRB: 9.44e-08    \n"
     ]
    }
   ],
   "source": [
    "data_with_too_much_signal = data.copy()\n",
    "\n",
    "data_with_too_much_signal[12:17] = np.random.normal(rand_mean, rand_std, size=data_with_too_much_signal.shape[1])\n",
    "run_model(data_with_too_much_signal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Explaining: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 50/50 [00:01<00:00, 37.55it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "heatmap_with_too_much_signal = run_explainer(data_with_too_much_signal)\n",
    "plot_data_and_heatmap(data_with_too_much_signal, heatmap_with_too_much_signal)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model now classifies this as noise. The vertical line in the center is still pushing it towards the FRB class, but not enough . We do see that the bad channels cause the overall input to be more classified as noise."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.1"
  },
  "vscode": {
   "interpreter": {
    "hash": "f74811edbe99894b2f930b63702daebe3ce5897f538d47d6f6827e4475af2be0"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}