STEP2b_MVPA_test_subcor.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/mnt/chrastil/lab/snlbin/programs/mnt/lib/python3.7/site-packages/sklearn/utils/deprecation.py:144: FutureWarning: The sklearn.linear_model.base module is  deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.linear_model. Anything that cannot be imported from sklearn.linear_model is now part of the private API.\n",
      "  warnings.warn(message, FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import os\n",
    "import glob\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from nltools.data import Brain_Data\n",
    "from nltools.mask import expand_mask\n",
    "from bids import BIDSLayout, BIDSValidator\n",
    "\n",
    "#data_dir = '/mnt/chrastil/lab/users/lily/localizer_beta_series_run_trials_only'\n",
    "#data_dir = '/mnt/chrastil/lab/users/lily/localizer_beta_series_all_run_exploration'\n",
    "#data_dir = '/mnt/chrastil/lab/users/lily/localizer_beta_series_all_exploration'\n",
    "#data_dir = '/mnt/chrastil/lab/users/lily/path_direction/localizer_beta_series_run_trials_only'\n",
    "# data_dir = '/mnt/chrastil/lab/users/lily/path_direction/localizer_beta_series_all_exploration'\n",
    "data_dir = '/mnt/chrastil/lab/users/lily/path_direction/localizer_beta_series_all_run_exploration'\n",
    "\n",
    "layout = BIDSLayout(data_dir, derivatives=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltools.utils import get_resource_path\n",
    "from nltools.file_reader import onsets_to_dm\n",
    "from nltools.data import Design_Matrix\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub_num = 'sub-043'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## All trials"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "E_file_list = glob.glob(os.path.join(data_dir, 'derivatives', 'nibetaseries',sub_num,'func','*Run*_desc-E*.nii.gz'))\n",
    "E_file_list.sort() \n",
    "E = Brain_Data(E_file_list)\n",
    "\n",
    "N_file_list = glob.glob(os.path.join(data_dir,  'derivatives', 'nibetaseries',sub_num,'func','*Run*_desc-N*.nii.gz'))\n",
    "N_file_list.sort()\n",
    "N = Brain_Data(N_file_list)\n",
    "\n",
    "W_file_list = glob.glob(os.path.join(data_dir, 'derivatives', 'nibetaseries',sub_num,'func','*Run*_desc-W*.nii.gz'))\n",
    "W_file_list.sort() \n",
    "W = Brain_Data(W_file_list)\n",
    "\n",
    "S_file_list = glob.glob(os.path.join(data_dir, 'derivatives', 'nibetaseries',sub_num,'func','*Run*_desc-S*.nii.gz'))\n",
    "S_file_list.sort() \n",
    "S = Brain_Data(S_file_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = E.append(N)\n",
    "data = data.append(W)\n",
    "data = data.append(S)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "109\n",
      "84\n",
      "74\n",
      "64\n"
     ]
    }
   ],
   "source": [
    "print(len(E.data))\n",
    "print(len(N.data))\n",
    "print(len(W.data))\n",
    "print(len(S.data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Y = pd.DataFrame(np.hstack([np.zeros(157), np.ones(119),2*np.ones(122), 3*np.ones(97)]))\n",
    "Y = pd.DataFrame(np.hstack([np.zeros(len(E.data)), np.ones(len(N.data)),2*np.ones(len(W.data)), 3*np.ones(len(S.data))]))\n",
    "\n",
    "#Y = pd.DataFrame(np.hstack([np.zeros(42), np.ones(44)]))\n",
    "\n",
    "data.Y = Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<bound method Brain_Data.shape of nltools.data.brain_data.Brain_Data(data=(331, 238955), Y=331, X=(0, 0), mask=MNI152_T1_2mm_brain_mask.nii.gz, output_file=[])>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Selection - Whole Brain Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ------------------\n",
    "# Define the prediction function to be used.\n",
    "# Here we use a Support Vector Classification, with a linear kernel\n",
    "from sklearn.svm import SVC\n",
    "svc = SVC(kernel='linear')\n",
    "\n",
    "# Define the dimension reduction to be used.\n",
    "# Here we use a classical univariate feature selection based on F-test,\n",
    "# namely Anova. When doing full-brain analysis, it is better to use\n",
    "# SelectPercentile, keeping 5% of voxels\n",
    "# (because it is independent of the resolution of the data).\n",
    "from sklearn.feature_selection import SelectPercentile, f_classif\n",
    "feature_selection = SelectPercentile(f_classif, percentile=5)\n",
    "\n",
    "# We have our classifier (SVC), our feature selection (SelectPercentile),and now,\n",
    "# we can plug them together in a *pipeline* that performs the two operations\n",
    "# successively:\n",
    "from sklearn.pipeline import Pipeline\n",
    "anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#############################################################################\n",
    "# Fit the decoder and predict\n",
    "# ----------------------------\n",
    "anova_svc.fit(data.data, data.Y.values.ravel())\n",
    "y_pred = anova_svc.predict(data.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification accuracy: 0.2296 / Chance level: 0.250000\n"
     ]
    }
   ],
   "source": [
    "#############################################################################\n",
    "# Obtain prediction scores via cross validation\n",
    "# -----------------------------------------------\n",
    "from sklearn.model_selection import LeaveOneGroupOut, cross_val_score\n",
    "\n",
    "# Define the cross-validation scheme used for validation.\n",
    "# Here we use a LeaveOneGroupOut cross-validation on the session group\n",
    "# which corresponds to a leave-one-session-out\n",
    "#cv = LeaveOneGroupOut()\n",
    "\n",
    "\n",
    "# Compute the prediction accuracy for the different folds (i.e. session)\n",
    "#cv_scores = cross_val_score(anova_svc, data.data, data.Y, cv=cv, groups=4)\n",
    "cv_scores = cross_val_score(anova_svc, data.data, data.Y.values.ravel(), cv=5)\n",
    "\n",
    "# Return the corresponding mean prediction accuracy\n",
    "classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# Print the results\n",
    "print(\"Classification accuracy: %.4f / Chance level: %f\" %\n",
    "      (classification_accuracy, 1. / len(data.Y[0].unique())))\n",
    "# Classification accuracy:  0.70370 / Chance level: 0.5000\n",
    "\n",
    "\n",
    "#############################################################################"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# roi - thalamus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "atlas_mni_file = os.path.join(data_dir,\n",
    "                              \"derivatives\",\n",
    "                              \"data\",\n",
    "                              \"HarvardOxford-sub-maxprob-thr25-1mm.nii.gz\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "threshold is ignored for simple axial plots\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mask = Brain_Data(atlas_mni_file)\n",
    "mask_x = expand_mask(mask)\n",
    "\n",
    "f = mask.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# index - 1 for all because this is python!!!\n",
    "tha = mask_x[[4-1,15-1]].sum() # thalamus\n",
    "\n",
    "data_masked = data.apply_mask(tha)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification accuracy: 0.2689 / Chandata_maskedce level: 0.250000\n"
     ]
    }
   ],
   "source": [
    "cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# Return the corresponding mean prediction accuracy\n",
    "classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# Print the results\n",
    "print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "      (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# new algorithms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import KFold, LeaveOneOut, GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data_masked.data\n",
    "Y = data_masked.Y.values.ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.3293051359516616\n"
     ]
    }
   ],
   "source": [
    "loo = LeaveOneOut()\n",
    "X_std = MinMaxScaler().fit_transform(X)\n",
    "svc = SVC(kernel = \"rbf\",C = 0.001, gamma = 0.001)\n",
    "\n",
    "cv_scores = cross_val_score(svc, X_std, Y, cv= loo, n_jobs = -1)\n",
    "\n",
    "classification_accuracy = cv_scores.mean()\n",
    "print(classification_accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import permutation_test_score\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "anova_svc = Pipeline([('anova', feature_selection), ('svc', SVC(kernel = \"linear\"))])\n",
    "\n",
    "null_cv_scores, value2, value3 = permutation_test_score(\n",
    "    anova_svc, X, Y, scoring=\"accuracy\", cv=loo,  n_jobs = -1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.24169184290030213\n",
      "[0.29305136 0.26888218 0.28096677 0.25981873 0.3141994  0.25679758\n",
      " 0.22054381 0.28398792 0.22054381 0.27492447 0.25981873 0.24471299\n",
      " 0.25981873 0.29607251 0.29003021 0.32628399 0.28096677 0.28700906\n",
      " 0.33534743 0.27190332 0.23867069 0.25679758 0.25981873 0.23867069\n",
      " 0.21752266 0.2326284  0.26283988 0.27794562 0.24773414 0.29003021\n",
      " 0.23564955 0.25377644 0.29003021 0.3021148  0.29607251 0.29607251\n",
      " 0.29305136 0.21148036 0.25377644 0.25981873 0.35649547 0.26283988\n",
      " 0.21752266 0.27794562 0.29003021 0.24773414 0.23867069 0.32024169\n",
      " 0.24773414 0.34743202 0.24471299 0.28398792 0.32326284 0.33232628\n",
      " 0.32024169 0.26283988 0.32930514 0.23564955 0.26283988 0.27794562\n",
      " 0.25679758 0.25075529 0.29305136 0.25075529 0.31117825 0.23564955\n",
      " 0.27492447 0.25377644 0.27794562 0.23564955 0.31117825 0.20241692\n",
      " 0.24169184 0.25679758 0.26283988 0.28096677 0.28398792 0.25981873\n",
      " 0.29909366 0.21148036 0.25075529 0.29607251 0.27492447 0.24169184\n",
      " 0.29003021 0.28096677 0.3021148  0.28096677 0.29305136 0.25377644\n",
      " 0.20543807 0.26283988 0.27794562 0.27190332 0.27190332 0.26283988\n",
      " 0.25679758 0.3081571  0.33232628 0.31117825]\n",
      "0.8415841584158416\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.27196374622356495"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(null_cv_scores)\n",
    "print(value2)\n",
    "print(value3)\n",
    "\n",
    "value2.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.hist(value2, bins=20, density=True)\n",
    "ax.axvline(null_cv_scores, ls='--', color='r')\n",
    "score_label = (f\"Score on original\\ndata: {null_cv_scores:.2f}\\n\"\n",
    "               f\"(p-value: {value3:.3f})\")\n",
    "ax.text(0.7, 260, score_label, fontsize=12)\n",
    "ax.set_xlabel(\"Accuracy score\")\n",
    "_ = ax.set_ylabel(\"Probability\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Retrosplenial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# atlas_mni_file = os.path.join(data_dir,\n",
    "#                               \"derivatives\",\n",
    "#                               \"data\",\n",
    "#                               \"Schaefer2018_100Parcels_17Networks_order_FSLMNI152_1mm.nii.gz\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# mask = Brain_Data(atlas_mni_file)\n",
    "# mask_x = expand_mask(mask)\n",
    "\n",
    "# f = mask.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "# rsc = mask_x[[48-1,96-1]].sum() # retrosplenial cortex\n",
    "\n",
    "# data_masked = data.apply_mask(rsc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Precuneus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "# pcun = mask_x[[35-1,36-1]].sum() # precuneus\n",
    "\n",
    "# data_masked = data.apply_mask(pcun)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Medial Parietal Lobe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "\n",
    "# mpar = mask_x[[24-1,73-1]].sum() # medial parietal lobe\n",
    "\n",
    "# data_masked = data.apply_mask(mpar)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Intraparietal Sulcus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "\n",
    "# ips = mask_x[[31-1,80-1]].sum() # intrapariental sulcus\n",
    "\n",
    "# data_masked = data.apply_mask(ips)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Extrastriate Cortex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "\n",
    "# ext = mask_x[[1-1,2-1,4-1,51-1,52-1,53-1]].sum() # extrastriate cortex\n",
    "\n",
    "# data_masked = data.apply_mask(ext)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Parahippocampus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "\n",
    "# phc = mask_x[[49-1,97-1]].sum() # parahippocampus\n",
    "\n",
    "\n",
    "# data_masked = data.apply_mask(phc)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Somatomotor Cortex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# index - 1 for all because this is python!!!\n",
    "\n",
    "# som = mask_x[[8-1,9-1,10-1,11-1,12-1,13-1,57-1,58-1,59-1,60-1,61-1,62-1,63-1,64-1]].sum() # somatomotor\n",
    "\n",
    "\n",
    "\n",
    "# data_masked = data.apply_mask(som)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROI - Temporal + Temporoparietal lobe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # index - 1 for all because this is python!!!\n",
    "\n",
    "# temp = mask_x[[41-1,42-1,50-1, 98-1,99-1,100-1]].sum() # temporal lobe, temporal parietal lobe\n",
    "\n",
    "# data_masked = data.apply_mask(temp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# anova_svc = Pipeline([('anova', feature_selection), ('svc', svc)])\n",
    "# anova_svc.fit(data_masked.data, data_masked.Y.values.ravel())\n",
    "# y_pred = anova_svc.predict(data_masked.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cv_scores = cross_val_score(anova_svc, data_masked.data, data_masked.Y.values.ravel(), cv=5)\n",
    "\n",
    "# # Return the corresponding mean prediction accuracy\n",
    "# classification_accuracy = cv_scores.mean()\n",
    "\n",
    "# # Print the results\n",
    "# print(\"Classification accuracy: %.4f / Chandata_maskedce level: %f\" %\n",
    "#       (classification_accuracy, 1. / len(data.Y[0].unique())))"
   ]
  }
 ],
 "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}