test_predict_image_category_with_MLP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torch.utils.data import DataLoader, TensorDataset\n",
    "\n",
    "import DeepMice.utils.data_loader as dl\n",
    "from DeepMice.models.nets import MLPNet\n",
    "from DeepMice.models.train import train_test_classification\n",
    "from DeepMice.utils.data_shuffle import shuffle_and_split_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Keys in the data dictionary:\n",
      " dict_keys(['neuron_fluoresence', 'neuron_activity', 'neuron_time', 'running_pandas', 'stim_details_pandas', 'pupil_area_pandas', 'licks_pandas', 'reward_pandas'])\n"
     ]
    }
   ],
   "source": [
    "# read neural data and labels\n",
    "data = dl.load_example_data(path='one_example_session.npy', print_keys=True)\n",
    "\n",
    "stimulus_details = data['stim_details_pandas']\n",
    "nr_frames_after=5 \n",
    "neuron_time = data['neuron_time']\n",
    "activity = data['neuron_activity']\n",
    "[trial_matrix_3d, image_index, is_change] = dl.get_trial_matrix_3d(activity, neuron_time, stimulus_details, nr_frames_after)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reshape (trial x feature) and preprocess\n",
    "nr_trials = trial_matrix_3d.shape[0]\n",
    "trial_matrix = trial_matrix_3d.reshape(nr_trials,-1) \n",
    "# normalize unit activations across trials\n",
    "mean = np.mean(trial_matrix, 0)\n",
    "std = np.std(trial_matrix, 0)\n",
    "trial_matrix = (trial_matrix - mean)/std\n",
    "\n",
    "# split data\n",
    "seed = 2021\n",
    "ratio = 0.2\n",
    "X_test, y_test, X_train, y_train = shuffle_and_split_data(trial_matrix, image_index, ratio, seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train_loader and test_loader\n",
    "batch_size = 128\n",
    "g_seed = torch.Generator()\n",
    "g_seed.manual_seed(seed)\n",
    "test_data = TensorDataset(torch.from_numpy(X_test), torch.from_numpy(y_test))\n",
    "test_loader = DataLoader(test_data,\n",
    "                            batch_size=batch_size,\n",
    "                            shuffle=False,\n",
    "                            num_workers=1,\n",
    "                            worker_init_fn=seed_worker,\n",
    "                            generator=g_seed)\n",
    "\n",
    "train_data = TensorDataset(torch.from_numpy(X_train), torch.from_numpy(y_train))\n",
    "train_loader = DataLoader(train_data,\n",
    "                             batch_size=batch_size,\n",
    "                             shuffle=True,\n",
    "                             num_workers=1,\n",
    "                             worker_init_fn=seed_worker,\n",
    "                             generator=g_seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set MLP achitecture\n",
    "input_feature_num = X_train.shape[1]\n",
    "hidden_unit_nums = [100, 100]\n",
    "output_feature_num = 9\n",
    "cat_net = MLPNet('ReLU()', input_feature_num, hidden_unit_nums, output_feature_num).to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'cuda'"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9191f8a80feb4022ad54cda944eda6cc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value=''), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "Accuracy on the 3841 training samples: 100.00\n",
      "Accuracy on the 960 testing samples: 87.92\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(100.0, 87.91666666666667)"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.Adam(cat_net.parameters(), lr=3e-4)\n",
    "train_test_classification(cat_net, criterion, optimizer, train_loader,\n",
    "                              test_loader, num_epochs=100, verbose=True,\n",
    "                              training_plot=True, device=device)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}