mikbuch/language_localizer-fitting_glm.ipynb

## language_localizer-fitting_glm.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.7.0'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import nilearn\n",
    "nilearn.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nilearn.datasets import fetch_language_localizer_demo_dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/mb/nilearn_data/fMRI-language-localizer-demo-dataset\n"
     ]
    }
   ],
   "source": [
    "data_dir, _ = fetch_language_localizer_demo_dataset()\n",
    "print(data_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model was calculated elsewhere. Here it is just a \"given\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/nilearn/glm/__init__.py:55: FutureWarning: The nilearn.glm module is experimental. It may change in any future release of Nilearn.\n",
      "  warn('The nilearn.glm module is experimental. '\n",
      "/usr/local/lib/python3.9/site-packages/nilearn/glm/first_level/first_level.py:849: UserWarning: SliceTimingRef not found in file /Users/mb/nilearn_data/fMRI-language-localizer-demo-dataset/derivatives/sub-01/func/sub-01_task-languagelocalizer_desc-preproc_bold.json. It will be assumed that the slice timing reference is 0.0 percent of the repetition time. If it is not the case it will need to be set manually in the generated list of models\n",
      "  warn('SliceTimingRef not found in file %s. It will be assumed'\n"
     ]
    }
   ],
   "source": [
    "from nilearn.glm.first_level import first_level_from_bids\n",
    "task_label = 'languagelocalizer'\n",
    "models, models_run_imgs, models_events, models_confounds = \\\n",
    "    first_level_from_bids(\n",
    "        data_dir, task_label,\n",
    "        img_filters=[('desc', 'preproc')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['sub-01_task-languagelocalizer_desc-preproc_bold.nii.gz']\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "print([os.path.basename(run) for run in models_run_imgs[0]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "subjects z_map language network (unc p<0.001)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 187.2x165.6 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(45, 53, 43)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-6-a9c339debf88>:14: DeprecationWarning: get_data() is deprecated in favor of get_fdata(), which has a more predictable return type. To obtain get_data() behavior going forward, use numpy.asanyarray(img.dataobj).\n",
      "\n",
      "* deprecated from version: 3.0\n",
      "* Will raise <class 'nibabel.deprecator.ExpiredDeprecationError'> as of version: 5.0\n",
      "  zmap_data = zmap.get_data()\n"
     ]
    }
   ],
   "source": [
    "from nilearn import plotting\n",
    "from scipy.stats import norm\n",
    "p001_unc = norm.isf(0.001)\n",
    "\n",
    "models[0].fit(models_run_imgs[0], models_events[0], models_confounds[0])\n",
    "\n",
    "zmap = models[0].compute_contrast('language-string')\n",
    "plotting.plot_glass_brain(zmap, colorbar=False, threshold=p001_unc,\n",
    "                          title=('language-string (sub 1)'),\n",
    "                          plot_abs=False, display_mode='x')\n",
    "print('subjects z_map language network (unc p<0.001)')\n",
    "plotting.show()\n",
    "\n",
    "zmap_data = zmap.get_data()\n",
    "print(zmap_data.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check for which voxel there is the highest fit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(array([10]), array([18]), array([19]))\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "idx_max = np.where(zmap_data == np.amax(zmap_data))\n",
    "print(idx_max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[11788.172  11687.29   11859.798  11796.619  11442.766  11812.862\n",
      " 11660.242  11986.182  11951.117  12091.203  12086.89   12044.28\n",
      " 11944.157  11700.44   11869.953  11882.64   11895.412  11708.982\n",
      " 11732.074  11867.364  11747.167  11703.02   11543.501  11760.28\n",
      " 11860.697  11926.914  11981.654  12206.901  11858.295  11806.349\n",
      " 11542.439  11624.101  11664.395  11413.899  11507.76   11687.341\n",
      " 11853.974  11939.826  11535.203  11724.581  11725.045  11659.641\n",
      " 11432.441  11676.437  11787.974  11918.539  11960.478  11804.134\n",
      " 11778.379  11869.862  11804.942  11856.98   11561.879  11656.696\n",
      " 11788.041  11785.115  11573.504  11764.532  11665.229  11527.323\n",
      " 11621.351  11575.175  11657.655  11805.273  11987.177  12005.525\n",
      " 11898.305  11727.755  11684.403  11783.469  11521.2    11581.071\n",
      " 11641.178  11488.303  11147.072  11365.801  11492.592  11624.74\n",
      " 11804.441  11678.146  11613.721  11561.13   11546.337  11833.377\n",
      " 11784.704  11991.077  11893.536  11774.294  11776.242  11616.397\n",
      " 11577.12   11418.967  11554.54   11637.512  11599.026  11536.56\n",
      " 11462.901  11646.255  11760.586  11446.771  11749.047  11631.508\n",
      " 11671.786  11783.661  11932.712  11848.248  11883.583  11745.954\n",
      " 11507.626  11547.292  11469.245  11454.98   11435.081  11483.09\n",
      " 11626.427  11687.747  11869.706  11553.684  11339.196  11645.479\n",
      " 11828.976  11888.042  11786.766  11770.644  12031.609  11746.182\n",
      " 11750.422  11609.598  11632.092  11595.507  11615.021  11405.623\n",
      " 11546.199  11534.24   11486.845  11425.303  11602.134  11327.761\n",
      " 11323.692  11476.804  11638.955  11561.002  11730.728  11479.235\n",
      " 11328.429  11613.605  11487.344  11585.904  11479.316  11567.285\n",
      " 11573.196  11537.836  11771.139  11764.712  11531.298  11575.137\n",
      " 11490.341  11505.033  11652.024  11726.86   11895.278  11762.761\n",
      " 11691.264  11645.747  11681.08   11554.405  11525.539  11759.124\n",
      " 11510.785  11621.784  11523.272  11528.882  11564.741  11601.493\n",
      " 11540.659  11729.849  11914.777  11784.222  11647.36   11785.424\n",
      " 11993.808  11813.881  11858.428  11651.76   11501.839  11430.673\n",
      " 11542.753  11428.376  11659.845  11467.191  11548.4    11592.25\n",
      " 11382.946  11868.839  11526.84   11530.042  11396.008  11613.588\n",
      " 11887.165  11970.497  11837.237  11675.362  11530.948  11475.398\n",
      " 11374.456  11528.962  11551.184  11607.784  11587.869  11533.436\n",
      " 11909.508  11530.342  11619.656  11446.391  11560.361  11836.811\n",
      " 11758.661  11782.509  12074.074  11794.769  11931.79   11967.356\n",
      " 11576.442  11647.863  11705.621  11586.242  11510.952  11671.1045\n",
      " 11565.567 ]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-8-afe252673b32>:5: DeprecationWarning: get_data() is deprecated in favor of get_fdata(), which has a more predictable return type. To obtain get_data() behavior going forward, use numpy.asanyarray(img.dataobj).\n",
      "\n",
      "* deprecated from version: 3.0\n",
      "* Will raise <class 'nibabel.deprecator.ExpiredDeprecationError'> as of version: 5.0\n",
      "  data =  img.get_data()\n"
     ]
    }
   ],
   "source": [
    "from nilearn import image\n",
    "image.load_img(models_run_imgs[0])\n",
    "\n",
    "img = image.load_img(models_run_imgs[0])\n",
    "data =  img.get_data()\n",
    "\n",
    "# Let `y` be our observed signal\n",
    "y = data[idx_max][0]\n",
    "\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we need a modeled signal timecourse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>language</th>\n",
       "      <th>string</th>\n",
       "      <th>RotX</th>\n",
       "      <th>RotY</th>\n",
       "      <th>RotZ</th>\n",
       "      <th>X</th>\n",
       "      <th>Y</th>\n",
       "      <th>Z</th>\n",
       "      <th>drift_1</th>\n",
       "      <th>drift_2</th>\n",
       "      <th>drift_3</th>\n",
       "      <th>drift_4</th>\n",
       "      <th>drift_5</th>\n",
       "      <th>drift_6</th>\n",
       "      <th>constant</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.049782</td>\n",
       "      <td>-0.015876</td>\n",
       "      <td>0.006005</td>\n",
       "      <td>-2.89331</td>\n",
       "      <td>7.02019</td>\n",
       "      <td>-5.27795</td>\n",
       "      <td>0.093452</td>\n",
       "      <td>0.093445</td>\n",
       "      <td>0.093434</td>\n",
       "      <td>0.093419</td>\n",
       "      <td>0.093399</td>\n",
       "      <td>0.093375</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.049579</td>\n",
       "      <td>-0.015300</td>\n",
       "      <td>0.005969</td>\n",
       "      <td>-2.89903</td>\n",
       "      <td>7.01277</td>\n",
       "      <td>-5.25286</td>\n",
       "      <td>0.093434</td>\n",
       "      <td>0.093375</td>\n",
       "      <td>0.093276</td>\n",
       "      <td>0.093137</td>\n",
       "      <td>0.092960</td>\n",
       "      <td>0.092742</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.049899</td>\n",
       "      <td>-0.015055</td>\n",
       "      <td>0.005724</td>\n",
       "      <td>-2.88839</td>\n",
       "      <td>7.03311</td>\n",
       "      <td>-5.26949</td>\n",
       "      <td>0.093399</td>\n",
       "      <td>0.093234</td>\n",
       "      <td>0.092960</td>\n",
       "      <td>0.092576</td>\n",
       "      <td>0.092083</td>\n",
       "      <td>0.091482</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.5</th>\n",
       "      <td>0.000177</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.050209</td>\n",
       "      <td>-0.015752</td>\n",
       "      <td>0.005528</td>\n",
       "      <td>-2.88713</td>\n",
       "      <td>7.02144</td>\n",
       "      <td>-5.27385</td>\n",
       "      <td>0.093346</td>\n",
       "      <td>0.093023</td>\n",
       "      <td>0.092486</td>\n",
       "      <td>0.091735</td>\n",
       "      <td>0.090774</td>\n",
       "      <td>0.089602</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>0.028586</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.049601</td>\n",
       "      <td>-0.015629</td>\n",
       "      <td>0.006192</td>\n",
       "      <td>-2.88762</td>\n",
       "      <td>7.01049</td>\n",
       "      <td>-5.24182</td>\n",
       "      <td>0.093276</td>\n",
       "      <td>0.092742</td>\n",
       "      <td>0.091856</td>\n",
       "      <td>0.090619</td>\n",
       "      <td>0.089037</td>\n",
       "      <td>0.087116</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>336.0</th>\n",
       "      <td>-0.234251</td>\n",
       "      <td>0.156768</td>\n",
       "      <td>0.036007</td>\n",
       "      <td>-0.016914</td>\n",
       "      <td>0.007147</td>\n",
       "      <td>-2.73960</td>\n",
       "      <td>6.80817</td>\n",
       "      <td>-4.95885</td>\n",
       "      <td>-0.093276</td>\n",
       "      <td>0.092742</td>\n",
       "      <td>-0.091856</td>\n",
       "      <td>0.090619</td>\n",
       "      <td>-0.089037</td>\n",
       "      <td>0.087116</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337.5</th>\n",
       "      <td>-0.221650</td>\n",
       "      <td>0.475391</td>\n",
       "      <td>0.035995</td>\n",
       "      <td>-0.017398</td>\n",
       "      <td>0.007032</td>\n",
       "      <td>-2.75871</td>\n",
       "      <td>6.85601</td>\n",
       "      <td>-5.02555</td>\n",
       "      <td>-0.093346</td>\n",
       "      <td>0.093023</td>\n",
       "      <td>-0.092486</td>\n",
       "      <td>0.091735</td>\n",
       "      <td>-0.090774</td>\n",
       "      <td>0.089602</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>339.0</th>\n",
       "      <td>-0.172587</td>\n",
       "      <td>0.842279</td>\n",
       "      <td>0.036602</td>\n",
       "      <td>-0.017619</td>\n",
       "      <td>0.006823</td>\n",
       "      <td>-2.77903</td>\n",
       "      <td>6.84041</td>\n",
       "      <td>-5.05199</td>\n",
       "      <td>-0.093399</td>\n",
       "      <td>0.093234</td>\n",
       "      <td>-0.092960</td>\n",
       "      <td>0.092576</td>\n",
       "      <td>-0.092083</td>\n",
       "      <td>0.091482</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>340.5</th>\n",
       "      <td>-0.118173</td>\n",
       "      <td>1.119359</td>\n",
       "      <td>0.037354</td>\n",
       "      <td>-0.017272</td>\n",
       "      <td>0.007427</td>\n",
       "      <td>-2.73828</td>\n",
       "      <td>6.84949</td>\n",
       "      <td>-5.05414</td>\n",
       "      <td>-0.093434</td>\n",
       "      <td>0.093375</td>\n",
       "      <td>-0.093276</td>\n",
       "      <td>0.093137</td>\n",
       "      <td>-0.092960</td>\n",
       "      <td>0.092742</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>342.0</th>\n",
       "      <td>-0.073405</td>\n",
       "      <td>1.204995</td>\n",
       "      <td>0.038029</td>\n",
       "      <td>-0.016605</td>\n",
       "      <td>0.007377</td>\n",
       "      <td>-2.74262</td>\n",
       "      <td>6.84819</td>\n",
       "      <td>-5.05538</td>\n",
       "      <td>-0.093452</td>\n",
       "      <td>0.093445</td>\n",
       "      <td>-0.093434</td>\n",
       "      <td>0.093419</td>\n",
       "      <td>-0.093399</td>\n",
       "      <td>0.093375</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>229 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       language    string      RotX      RotY      RotZ        X        Y  \\\n",
       "0.0    0.000000  0.000000  0.049782 -0.015876  0.006005 -2.89331  7.02019   \n",
       "1.5    0.000000  0.000000  0.049579 -0.015300  0.005969 -2.89903  7.01277   \n",
       "3.0    0.000000  0.000000  0.049899 -0.015055  0.005724 -2.88839  7.03311   \n",
       "4.5    0.000177  0.000000  0.050209 -0.015752  0.005528 -2.88713  7.02144   \n",
       "6.0    0.028586  0.000000  0.049601 -0.015629  0.006192 -2.88762  7.01049   \n",
       "...         ...       ...       ...       ...       ...      ...      ...   \n",
       "336.0 -0.234251  0.156768  0.036007 -0.016914  0.007147 -2.73960  6.80817   \n",
       "337.5 -0.221650  0.475391  0.035995 -0.017398  0.007032 -2.75871  6.85601   \n",
       "339.0 -0.172587  0.842279  0.036602 -0.017619  0.006823 -2.77903  6.84041   \n",
       "340.5 -0.118173  1.119359  0.037354 -0.017272  0.007427 -2.73828  6.84949   \n",
       "342.0 -0.073405  1.204995  0.038029 -0.016605  0.007377 -2.74262  6.84819   \n",
       "\n",
       "             Z   drift_1   drift_2   drift_3   drift_4   drift_5   drift_6  \\\n",
       "0.0   -5.27795  0.093452  0.093445  0.093434  0.093419  0.093399  0.093375   \n",
       "1.5   -5.25286  0.093434  0.093375  0.093276  0.093137  0.092960  0.092742   \n",
       "3.0   -5.26949  0.093399  0.093234  0.092960  0.092576  0.092083  0.091482   \n",
       "4.5   -5.27385  0.093346  0.093023  0.092486  0.091735  0.090774  0.089602   \n",
       "6.0   -5.24182  0.093276  0.092742  0.091856  0.090619  0.089037  0.087116   \n",
       "...        ...       ...       ...       ...       ...       ...       ...   \n",
       "336.0 -4.95885 -0.093276  0.092742 -0.091856  0.090619 -0.089037  0.087116   \n",
       "337.5 -5.02555 -0.093346  0.093023 -0.092486  0.091735 -0.090774  0.089602   \n",
       "339.0 -5.05199 -0.093399  0.093234 -0.092960  0.092576 -0.092083  0.091482   \n",
       "340.5 -5.05414 -0.093434  0.093375 -0.093276  0.093137 -0.092960  0.092742   \n",
       "342.0 -5.05538 -0.093452  0.093445 -0.093434  0.093419 -0.093399  0.093375   \n",
       "\n",
       "       constant  \n",
       "0.0         1.0  \n",
       "1.5         1.0  \n",
       "3.0         1.0  \n",
       "4.5         1.0  \n",
       "6.0         1.0  \n",
       "...         ...  \n",
       "336.0       1.0  \n",
       "337.5       1.0  \n",
       "339.0       1.0  \n",
       "340.5       1.0  \n",
       "342.0       1.0  \n",
       "\n",
       "[229 rows x 15 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = models[0].design_matrices_[0]\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x12f6b0cd0>]"
      ]
     },
     "execution_count": 10,
     "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": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(model['language'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will use modeled \"language\" and \"string\" timecourses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 229)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "language = model['language'].values\n",
    "string = model['string'].values\n",
    "X = np.asmatrix(np.array([np.ones(len(y)),\n",
    "                          language,\n",
    "                          string]))\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(X, aspect='auto')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[11615.50603902   227.62084686    18.74765173]\n"
     ]
    }
   ],
   "source": [
    "betas = np.asarray(y * X.T * (X*X.T).I)[0]\n",
    "print(betas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABagAAACPCAYAAAAftTx4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAABQn0lEQVR4nO3de3zU1Z0//td7JpNkuJjIRZGAStXiqrgiwboN67dqFS0qFKvSutXdtav9uvXS/S4Wdl1Nrd8lymNXxa7dWrTW/uoFq1y6sV90RetKqyUI9dJKvReCyiUEkASSzJzfHzMJmfmcM/mcfC7zmcnr+XjwAD6ZzHxm3nPO55z3uXxEKQUiIiIiIiIiIiIiorDFin0CRERERERERERERDQ0MUFNREREREREREREREXBBDURERERERERERERFQUT1ERERERERERERERUFExQExEREREREREREVFRMEFNREREREREREREREUxYIJaRB4UkW0i8ka/Y4tF5C0ReU1ElotIbfb4OSKyXkRez/59Vr/feUFENonIxuyfw7LHq0TkcRF5R0ReEZGj/X+bRERERERERERERBQ1bmZQPwTgvLxjzwI4SSl1MoA/AliYPb4DwIVKqSkArgTw07zfu1wpdUr2z7bssasA7FJKHQvgLgB32L8NIiIiIiIiIiIiIio1AyaolVIvAmjLO/aMUqon+9+XAUzIHt+glNqaPf4mgKSIVA3wErMB/CT7758DOFtExOX5ExEREREREREREVGJqvDhOf4WwOOa4xcDeFUpdaDfsR+LSArAkwBuV0opAHUANgOAUqpHRHYDGI3MbGyjMWPGqKOPPtqH0yciIiIiIiIiIiKioKxfv36HUmqs7meeEtQi8s8AegD8LO/4ichs1XFuv8OXK6VaRWQkMgnqrwN42PL1rgZwNQAceeSRaGlp8XD2RERERERERERERBQ0EfnQ9DM3e1CbnvSvAVyATOJZ9Ts+AcByAFcopd7tPa6Uas3+vRfAIwBOy/6oFcDE7O9WAKgBsFP3mkqp+5VS9Uqp+rFjtQl3IhqEFRta0dC0BpMWNKOhaQ1WbGgt9ikREREREREREdEQMKgEtYicB+AmABcppTr6Ha8F0AxggVJqbb/jFSIyJvvvBDKJ7TeyP16FzA0VAeArANb0T3gTUbBWbGjFwqdeR2t7JxSA1vZOLHzqdSapiYiIiIiIiIgocAMmqEXkUQC/ATBZRLaIyFUAvg9gJIBnRWSjiPxn9uHfAnAsgFuyxzeKyGEAqgCsFpHXAGxEZtb0j7K/8wCA0SLyDoB/ALDAv7dHRANZvHoTOrtTOcc6u1NYvHpTkc6IiIiIiIiIiIiGigH3oFZKfVVz+AHDY28HcLvhqaYZfmc/gEsGOg8iCsbW9k6r40RERERERERERH4Z9B7URFQextcmrY4TERERERERERH5hQlqoiFu/szJSCbiOceSiTjmz5xcpDMiIiIiIiIiIqKhYsAtPoiovM2ZWgcgsxf11vZOjK9NYv7MyX3HiYiIiIiIiIiIgsIENRFhztQ6JqSJiIiIiIiIiCh03OKDiIiIiIiIiIiIiIqCCWoiIiIiIiIiIiIiKgomqImIiIiIiIiIiIioKLgHNYVuxYZW7Q35TMeJiIiIiIiIiIioPDFBTaFasaEVC596HZ3dKQBAa3snFj71Olo+bMOT61sdxwEwSU1EREREREREVMY4aXFo4xYfFKrFqzf1JaF7dXan8Ogrm7XHF6/eFObpERERERERERFRiHonM7a2d0Lh4KTFFRtai31qFBImqClUW9s7tcdTSlk9noiIiIiIiIiISp9pMiMnLQ4d3OKDQjW+NolWTdI5LqJNUo+vTXKZBxERERERERFRmTJNTuSkxaGDM6gpVPNnTkYyEc85lkzE8dXPTdQeP/P4sVzmQURERFTiVmxoRUPTGkxa0IyGpjVsyxEREVGf8bVJq+NUfpigplDNmVqHRXOnoK42CQFQV5vEorlTcPucKdrjz7+1ncs8iIiIiEoY95UkIiKiQkyTGefPnFykM6KwiTLs/Rt19fX1qqWlpdinQQGbtKAZum+oAHi/aVbYp0NERERElhqa1mi3eKurTWLtgrOKcEZEREQUNdzetfyJyHqlVL3uZ672oBaRBwFcAGCbUuqk7LHFAC4E0AXgXQB/o5Rqz/5sIYCrAKQAXK+UWp09fh6AewDEASxVSjVlj08C8BiA0QDWA/i6UqprUO+Wyoppz2ou8yCiKGKjiojIiftKEhER0UDmTK1j32kIc7vFx0MAzss79iyAk5RSJwP4I4CFACAiJwCYB+DE7O/cJyJxEYkD+A8A5wM4AcBXs48FgDsA3KWUOhbALmSS20Rc5kFEJYNL2Esb98clCg73lSSicsY2BBGRd65mUCulXhSRo/OOPdPvvy8D+Er237MBPKaUOgDgfRF5B8Bp2Z+9o5R6DwBE5DEAs0XkDwDOAvC17GN+AqARwA+s3w0VTVCzBnufgzMS/cHZnUT+0JWlxas3GffMZzmLtt7Bhd749Q4uAGDsiHwwf+bknDIGcMIBEZUHtiGIiPzhKkHtwt8CeDz77zpkEta9tmSPAcDmvOOfQ2Zbj3alVI/m8TlE5GoAVwPAkUce6cuJk3dBX5S5zMMfbDwR+cNUlvKT0724hD36OLhAFCxOOCCicsU2RGngRK2hg7EuXZ4T1CLyzwB6APzM++kUppS6H8D9QOYmiUG/HrnDi3JpYJxKGy+00WEqS3ERpDQ3HuYS9ujj/rhEweOEAyIqR2xDRB8nag0djHVpc7sHtZaI/DUyN0+8XKm+XnkrgIn9HjYhe8x0fCeAWhGpyDtOJaLQRZn7cUUHG0+li3sbR4upzKSU4p75JYr74xIREdFgsA0RfYUmalF5YaxL26AT1CJyHoCbAFyklOro96NVAOaJSJWITAJwHIDfAlgH4DgRmSQilcjcSHFVNrH9PA7uYX0lgJWDPS8Kn+niW5NMBJZUY+Lb3lBtPJXDd4UX2mgxlZm62iQWzZ2CutokpN//OVoffbwhLxEREQ0G2xDRx4laQwdjXdpcbfEhIo8C+AKAMSKyBcCtABYCqALwrIgAwMtKqW8qpd4UkWUAfo/M1h9/r5RKZZ/nWwBWA4gDeFAp9Wb2Jb4D4DERuR3ABgAP+PT+KASmG9+IIJAtJbhsY2C67SCG4g2KyuW7wgtttBQqS1zCXpq4P27p4HZHREQUJWxDRN/42iRaNf2mcp+oVe50bULGurSJ0uyXWQrq6+tVS0tLsU+DsnSVw7cf3wjdt0sAvN80a9Cv1dC0Rlvp1NUmsXbBWYN+3nKRn5QFMsmzRXOnABhajadS/a7kl6eOrh7s6uh2PC7q76OcMUlGFL5C1zeWPyIiKoRtt6GL7YfyY4rpxdPq8OT6VsY6wkRkvVKqXvczzzdJJAL0N75ZvHpTIKNXnE1aWKHtINYuOGtIVcyl+F3RzfpOxASJuKA7dXDIp9xnv0cdZ0oThY83+yUiosEol1WVNDhDdZZ7OQ/KmNqEz7+1HYvmTinb913umKCmwAS1pUSQyzbKoRIvxaRsUEpxiY/uYtudVqhNJjC8qqKkv5tDVTnUK0MZ4xcdvL4R+ei1ZcBztwG7twA1E4CzbwFOvrTYZzVk8VoTLA5w0lCbXFLugzKF2oRDLdblhAlqCoxxpDK+Frhr8A1ivxLf+Q3BM48fm7McpFQrcd+SsmXQcSnFfbdNF9vdnd3YeOu5zh+UQZzK2YoNrXhp+X14HI9hfNUObO0Yg7uXzwNwbUnVK0OCpiytSDWUdeO+1IyvTWLanmdxU8UyjJcd2KrG4M6eS7H+kHO0j2fCxz/WnyWvTdH22jLgF9cD3dk2x+7Nmf8DjFMRFEwkxdeyLPkg8AFO1nmRor1mDbGyVLKDMi7LUilORKOBMUFNgZoTX4s5VbcB1VuAqglA67nA7x7x1CC2XaKju0ABcDQEf/bynxx7ZpdEJZ5nUEnZ/AvBcd7jFAWluJzLdLG9csRvgbu+k3uxBqw6mEzWhCCvLKX2nIjbZA2GSRcAYILswG3qftzZXIE5U79b5JMdwlzWeRvVNejsPi3nV0vxulAu7j7hbZy0fimS/crTHYmleOOEowHk7sdf7jOHgPDqdOvPksnP6HvutoPx6dXdmTnOGIXOlEja2Hw/5sgPWZZ8EGgyi3VepOiuWS8tvw8XJJaiIrU/86D+MQLKMnFdkqvOTGXpTy8Dbz+TE6P5MxtKbiIaDYw3SSxTkUhE5VcwADK3SNR852omAt9+w/dTMG2eX52IaW86p2N7U8ewP3tTAt71OUQgTnSQ7jv7lcpfo6l/owoAEkmgIgl0tjmfRBMn3hwkBJqylFZATJwP3ZIegwm3vRviyVEfizpvS3oMZnQtcRz3erNfckE3g+a52zKdlXyaOq9Ub5LrVph1esHP8ks7PMWJQpJfnnTxAQAI0Nge5pkRgEkLmrU3ln+p8npMiO1w/iA5CqgcXnYJtSAFWmfedZK+TDFOfcLsH+uuWQXLUk9nbpswkQQuXFLysSrJdpCpLOW307MxWpFqKH7Oi6zxJolDTGRmDelmZ2ibX8hcuANgmpGQf6wQm5H1sD970+stmjvF/YUnAnHySyQGZjzSzfq+TZ5ERef+3Ad2d2rilqWJk+0yr3L4LEOnKUu65DQAjI/tDOGESMuizjPFicsHA2aaQWOo89Tuzfik8VgcprZjm4zF5lPnY2v7BO1jIz1zyEKYS3dNn1n9nmeBX/zYdZywe3O288lkTah05ck4EWECtyooAtP2Rca2QmfbwQkKnK3rim+rKnXlw9Q/MsRpqCXVwu4fb23vxEWxl3LKU51oktOAfqJPmawmKcWtLs25hrzrVTZGc779RlmXnaGICeoyFJn9hmySmTX6jqRXth3R/Oa6bSUe9mfvy+tZxKkjOQ7DbE4wRJEZmPGB48YOjR/bPYGmPNks8yqnzzJUFmVpf3IchjEJUByWcfpK+te4EY/1dXLuxjzMmHltgCdIxu0HJA4o5wCzUsA42Q4IMA7bUbP+Zsyr/iYe3X+647HlMrgQ5tJd09L4hZVPWMUJkIMzo5hUC49xUE4zI+24c7lVQRGYti/qTtSgqrt94Ccok4Ra0DzfOM00eJo8VJ/ozNfdiY5f3oKFn949pNrYYfePrxzxW9zUvTRne720ytR4rkV8UpYbkd/qUtcPKrjCJ08ZxIicYsU+AfJfZPYbMiad8y4PiWSmQnptWWZmTWNt5u/Xlnk+BVNHtDaZQDIRzzmWTMRx+elHoq42CUFm+Yvtsq+wP3tfXs8Qp3TeQGWHqsSd3ZcFEic/FGr8lDxTWUqOypSf/no7mHkxMpUF3fGy/iyDZIhT/hy1nng1hp34pUynZvfmzCN6OzkRKU/lYsWGVjQ0rcGkBc1oaFqDFRtara5Nw078EpoSSzEhtgMxASbEdqApsTRzox0KjqnToVKOOi8N50qFpHThOjyivc5HeuaQBZs63VZ+uTnz+LHaz/JwGGakaeKknbHbm1SjYBWakVYzEYBk/r5wSWaPT9Pe1BSY6e/e25ec7pWULlRVxDRlyYDJmuCZBk8B13Gq7vx4yLWxw+4f35R4vC853SsmmjUjiSQOJGq1z9GRHBfIuYVtztQ6rF1wFt5vmoW1C84KPDmtbXfr9A725PeDjjvX0H7QCGiCIxUXE9RlKMhOi5EuaXn2LfrkWf3fOhvEQCDJmvkzJ2s7VY0XnYhFc6c4ktG3z5niqRIP+7O3fj2XcepQlfhp6ovYkh6DtBJsSY/Bgu5vYFdHV2STapEZmPFDfpx0F+tEEjj/jkz56V+e/vxrmZu95cXo7hPedp2sKavPMkyGOk/qr8qJUcXse5kECEHvSoDW9k4oHJyltO6Y69xfm95+JnfvdyDzf8YpWKZOR29c+sVJDDtSjVM7tdf5yMwc8sjUvvGagNeVmyfXt+LiaXWOz1Is4lSqW4eVhUJx+vYbmT2nv/1GZvatKR6MU7CMW0Tscpal5Cj9Y5msCZ4PcdqaHq0/XsZt7LD7x8M69StPBXC08xapv0aHqsx5XN+kLLJiandrk9SmwZ63n3GWpfq/1bfbz74lqLdCRcQtPsqQX/sNud6D1rTc6cIlmT9ulrDfdVIgdxMfaGmL3x3VsPd6sno9izjdue9iPHTgNNya9xQvV98Q2bu+B3p37jDp4vS7RzKJ57y7F/d95v0/e0NZmv7uvVg0d7WrMl02n6VPtHVhfK25bnNT5z11tf7FmATwzeLVm3BO6le4qTJ3T88bf38O1rq9NjFOxXH2Lc69jHs7IydfmhOrTxqPxThsdzzFNhnjfTl3hAW1dNe0gub5t7Y7720Rdx8n442PmFQLXqHylM+0vJpxClahzz2/LOlu9MtkTTgs49Sz8rqcQe6eeDWWVvwV0OV8inJuY4e+F7IxTs4b9f7kkWa0xbqy+1XvxFY1Gnf2XIpfHDgNjcGcXdmy2sql0GBoflkCgCNP57aIQwQT1GWoUKfFbdLZag9a0wjYc7cdnJExkABnbITZQQ17ryer17OI0ykbWpHUNCRMy3nV7i12+3oFoCRvBKFTaEQ5r1EFOJOnL+03xGL3Fm1Z0NUJZfNZ+kBXF760/D5ckFh6sNORv0enmzpviCYBwrz5Zv2eZ7EokbsHYVNiKRbuAXDyIsYpSnT7ELocRNh86nzUrL85Z3l8p6rE5mnzUR4LdM2CaN8UvCHiXdcPflDOJklK/mKcos/mc7eJJ/nLIk4rUg14qfsb2XtYZBKfd6fnofrUryC5vnVItbFD3wvZIk7ja5NY1T4Dq7pm5ByvK+MBg6BYrcC1bF+vSDVg8YEl2Lq/E+Ork5ifmow5AG/qW4ZEKcOSu4irr69XLS0txT6NUHnt2OcnWoDMBVG35LWhaY12BmVdbdI5g6axFvqlm5JZNuiGcWaNc6STBskyTrrv2+kr/5d2ptrHGItxje/4fcbWwkx++UF7vitPhNs46cr02qrr9Xeq1pSlQnUC4L0hqXt/fjxvmHR14UuV12NCzN1nbGSaAXXhkrJtWNlcg/zwsWFm7ccYi5dn/2pwK4SAso9T6Ayf8bop38WNvz/OVV2xbtUPMfHVxThM7cA2GYPNp87H9IuuCekNlBddnXdR7CXcUfkAkjhw8GC2HKxINbiv09mRDJ4fnzHjFDzdZwzwcy8Sm/6D2+tNob703Se8nX2O7dgmY3nNCoLLeszUNn14+oeY/u69LI8WrPJHFu3rgjF6/Va20UuQiKxXStVrf8YEdWnwo2NvU2lMWtBsSpHh/aZZuQf9SC4zCRA8H+J0wz8tzJmRCGT26VrY/Q3c86+L/DrTIcFUptePuBHDOj9y/oImTuZEwlIk+60f7IlXZ/Y9zitLVg0JS7r3l4gJIEB36mDtEmSC0g+6uvC9qq85bsqWYTEoB1glAUpt8EUnyO+bjmqshWiuZAqCE1KPub+eauJkSsqVQ5xCZ7g2taoxaDiwpO//Ua8ryoXNwGdH8ghM+/Tu0AadaABsS5eGIOPEwQVrNgnKFakG1/1xU196duwl3DP8xyynfvHhO5/fdrv7hLeZ+BwE23yV18Gel6tv0E5E4QTH6CuUoOYWHyXCak8fA5tlF1Z70PqxHJDL1YLnQ5xaDjkHC/bAsU/X+kPOCeCEy5upTN/ZfRkaEz90FSdd2V2VngF05cbo7vQ8zEg1ZJZCDfD7hY7b0L2/7rSzqV6oHotCsk9XF25VYzBBO0vdctsHl9uBWG25FGFh33xTDMsHP8EYu+tpXpxM8Wj5sA1P9lu2W6pxCp1hK68jsDPn/7ZtHhoc3VLs8ft3ah9b3fmxsSzlPwdnVoeg0FZuQzDxGdlVXJZx8nxPICDysSomXXv1nNSvcNKrDwC9q0ayn+VGdQ06u0/Leazp2mTqSy+sfCKy9/MpOT595x3bZd11PWM0CAW3csm7hqw75jpcse4odHbf0/f7yXVxLJrY6ihLpn7CYWo7tPta8j4xJW3ABLWIPAjgAgDblFInZY9dAqARwJ8BOE0p1ZI9fjmA+f1+/WQApyqlNorICwCOAND7DTtXKbVNRKoAPAxgGoCdAC5TSn3g/a2VFz869jZJZ6s9aP1KLrvdu3WI8pys8yFOme9FV84+XclEHIvKeN+0oJjK7k8+PQ2NXzvRVZxMZXpV2rmX2m8sGs9+3KjFpm7SPTYqSVldXXg35qEpvjTnxjdB7tHpxwBlFIR+803DoNyifZdoH+72O2u6+eKjr/wlUnmr0koxTiaBDRgZBhK2qtHOY7aDGSWYVIsCZ2fdEKO0M0bAwfraVf3NpJp/grqfSwnGSNeGmP/E73JWcRVtENEiTr7dE8hjnKIwYSAouuvKTRXLcrc0AoDuTnwj/f/hIZzmeLzuOUx9adP9fJhUG4SgvvMB3hur3GnvjZF/w9Ddm3HSqzfjnNQ3sAoH+6u2gz3bZKxhBjXvE1PKYi4e8xCA8/KOvQFgLoAX+x9USv1MKXWKUuoUAF8H8L5SamO/h1ze+3Ol1LbssasA7FJKHQvgLgB3WL+LIcDUgbfp2M+fORnJRDznmCnpPGdqHRbNnYK62iQEmSXYBZdsnnxpZilFY7v7GyMOMSs2tKKhaQ0mLWhGQ9MarNjQavW7C596Ha3tnVA42EC1eQ4AnuNk/b0go4Jl2mWcdGXaxNR4dlsn2LKpm3SPLZSUDZPuOz/jy9dmtkypmQhAMn/7uezvtWWZbQ8aa4G7TsrcnEwjqJnHQQny+6Z18qWZuOTFqcWw4sPtd7Z+z7NoSizFhNgOxASYEMvcfHGW/I/28aUWJ53BXINcX/POviUzwNNPJ6pwZ49+UM613qTa7s0A1MGk2mvLPF2PhyRNjJBIYmnlX2kfHhdxX38bEgwdv7yFMbJl6pR77awXSgJFlGkVV/8txoDitCts4mTVFgooqeZbHySidNeV8bpVcgDGx/SrSXTPYeozSVDldCgKKpHMGPmq45e35E7qAZBEF26qWOZ4rE1/dfOp87VtE97Ut7QNOINaKfWiiBydd+wPACCi3YSz11cBPObiHGYjMxsbAH4O4PsiIqpUN8cOiNWMZgPbO+ga7w4f9oykiM6AsplN4HU2qPUMygA/M9PIaBRjFGXzZ07GS8vvy97dOzML827Mw4yZ17p+Dl2Z7ujqwa6Obsdjx9cmHXGac/YtwFyLG1xZvj+3e1Dr6rGwt4MoRF8XBrTiQzNTranyAaiu7PYt/QQ28zggod/FHdCuzJmf0u+R5/Z6urDyCQzrt8c7AAyTLnwnsQyrDsxwPL7U4qRjew2yuuZpVve8ccx1eHbdUUB68G2eQonPhf32TeZWLC4YVmCdkmpAUlOW8r8rvXT1t9q9RbtCt7rjY7QeyDyeMXLJcis31+3YEpxN6HUVV6As4mTVFjKsRvGaVCuXVVwmuvbqRxiDOs1M5/3JcUim4q7bD9r2Y9yHrTEpI6DvvC/bl1Kf6s6PtcfHi3PAxzTYAzj7D9OnngccfShzEGUmyD2oL0Mm+dzfj0UkBeBJALdnk9B1ADYDgFKqR0R2AxgNmNa/DE1+deyNSWe3wl7mF9FlhbYJZ6+NO6sGKmM0aKbOWhBLG+fE1+KCxMFtIibIjsy2EfE/BzD4fdNMN6i4+4S3gV/c6ojTnAuXYM4C/+NkqrN0x3SfZejbQRQS0ACM9nv1gjOplsQBR/Iz0JnHAfJ8DTKxiJHX66lpee542elIzpVqnPLZDhhZX/PyBhKmA1g00WO9a0ieFdo3uRySLYHRDPbMyf6dH6fFqze5rr8/wRjtEt38LV4YIxcstnKzascGlQQKkKkNYXpsqE6+FOs+2JV7c7Ap8zHdYiu3wO4JpBGlCQM23LbddW2CrSfchDrNTfKGnX8bFqWmFH3LRcoKKpHMGPlqa3o0JsScbef867z1YA/ALWLLUCAJahH5HIAOpVT/22derpRqFZGRyCSov47M3tM2z3s1gKsB4Mgjj/TrdEtGYB17G0HegCUKr2eQ38jp6Oqx6uB6bdxZNVCHaIy8Cv3GZ8/d5ljuVJHa7/lzM44yv/CPocfJVGe5+dz8WDXii4AGYEzft9lx/WzC8bITdbXJstz/0bNBxMjL9dR080WpmYBFX/DYcY0o2wEj07Wtfs+zmZsPuRxI8PTZmfa2NuybHPVkS6gsB3x0cXJbfy/qugSLEksxTA6uSuhQldotXhgjF1x21q0GkUpwNqHXVVxBWrGhFQtd3hysKPcEyhOpCQMu2U4kctZjZxlnZs4xPIcVJtX8EWQimTHyzdLKv8JN3fc5rvP/Efsa+zbkENQM6nkAHu1/QCnVmv17r4g8AuA0ZBLUrQAmAtgiIhUAagA45/tnfvd+APcDQH19PbcAKYawl/lZvl4QM111jRwTU+fJ1LirSSbQ0LRmwPO1aqBGPEZRZeqsPfrK5mBufBbg56ZNGKwsrTgVZTsInYAGYEzft0/i+tmEUjMBa7991qBfr6yFPUhWIFkz5+QIDCQHwHbASHfNuyj2EpoqHwB2Z288FfRqG0Oclqq/Qt4OLX3nTPBlUM6m/m455Bws2JO5Kdl42YmtajTu7LnUsaURwBg5eFjdYzVxogRnE3pdxeUbTYwWrx7jenDAui0UQFItMhMGLPiyLQkTlKWBcYq8U2ZdjVuW9+BG9Vjfdf5uzMOMOd/EoqDqX4+rX8v5xrBR53uCWkRiyKxP/8t+xyoA1CqldohIAsAFAP47++NVAK4E8BsAXwGwhvtP2wm1AIW9zM/i9bzu82yia+SYmDpPppkc+7p60N7ZPeD5WjVQIxwjvwTxnTd11vKT0wM93rUhECevIrFqJKCBBNP3Z1HXJbhn+I9LaqZa0YU9SFaCyRqvbJMkumvedxLLkMSB3AcGOZBguW9ylJMtobbzfBrwcVt/Z74rXVjVdTAhnYgJEvHiz3SNNI8DCdazYkswCeRlFZcvDDGq3/c3aIVzAMbULtC9jzDrhMhMGLBgvXI1CvfSicI5RF0UPqMonEOJydQV1+Ky1WeHU4d4vD4GlVMidwZMUIvIowC+AGCMiGwBcCuANgD3AhgLoFlENiqlZmZ/5QwAm5VS7/V7mioAq7PJ6TgyyekfZX/2AICfisg72eed5/ldDSGhF6Cwl/lZvF5QN/Fwm4gcaN+k3nMsdDO7QufrOlkX4Rj5IajvvKmzFhfRJqltZ3LldybuPuE6TNfsb1cucSobASX2Td+3lkPOwboTjna1N2U58dTZLsbgSwkma7yyGTDSXfPG79cujgt2FYfFvslR7XSE3s4LecAnMjNdS43HgQTfZsUyWWNmiNHCyiewcv/gVwgMpk5Yt+qH2XbFdmyTsdh86nxMv+ga128lEhMGLFgNwEThXjpROIeoi8JnFIVziDrDNSHUOsTj9bHcbwwbdQMmqJVSXzX8aLnh8S8AOD3v2D4A0wyP3w/gkoHOg/RCL0BhzxyzeL2gbuJhauTUJhMYXlXhuvOUXzFPWtAcyPlGOUZ+COo7b+qsXTytLmcP6t7jNp04XWfiinVH4eHp38X0d+8tyziVjYAS+6bv25nHj8UV67pc7U1ZLjwn4Dj4EkmOzshd5oGEsJdSzomvxZyq24DqLUDVBCB+C2xuThum0Nt5RRjw0XZcX1tWMjEqCo8DCb7MimWypjBDLA7HDk831LWtE9at+iFOWn8zktIFCDAO21Gz/masA6yS1KXEagAmCvfSicI5RF3In5HbG5kzTv1E5Zrg8froV06J24QMTlB7UFNIAr2zsmEEbEWqAYsPLMHW/Z0YX53E/NTkvhlJgXA5Uy2om3iYGjmNF50YyIxd6/M1zV4J80Jg8XpeK+ugvvOFOmv1R43Szu5ys39473Oek/oVbqpchvGyA1vVGNzZcylu/P05WLvgDe3vBCKg70XYF+CgXk//vMEk9k3fN1PHc2Pz/ZjzwpNlObjgOQEXpcEXziY0MwwkrDvmunBnCEelA+VSoO08nSgM+JRYjIrCh4EEzzPamFQrLKAb6trWCRNfXZxJTveTlC5MfHUxUKYJaqsBmCjcS6cI51ByybMQPyPbG5lH9X4+oYvKNcHj9dGPHA23CRk8JqhLXGB3VjZ0DtZ9sCt75+kiFzZNEmD+zIZAbuIR1N5rviyvjHonLi9O6465zvP3J8i7iRfar7D/cduLTv2eZ7EosbTv7sUTZAeaEkuxcA8AFPnGdz7cRCLMC3BQr7diQyteWn4fHsdjGF+1A1s7xuDu5fMAXJtJUgdQnnTft28/vtHxuItiL+Gm7qXA7mwHM2rl3CPf9oos9mcR9frYgi8dV12cLlziOHbj02PQmdehCXSGcFQ6UC4Fec0Lc1DOSonFqCiiMJAQhcRelJ19C3pWXoeK1P6+Qz3xalR4vKFuoTpBV6YvUtuhy6odpnYM6vVLhesBGEMyqyM5Due4nIjiWcgrV0oyeRbiZ2R7I/Mo388nVFG5JlheH/PrzTOPH+t59TS3CRk8JqhLXGB3VjZ0Dia+ujhn6TlQhMJmSALMuXAJMLchkNHgRM1GDD/2Hozc9zGGDx+HRM0NaH5vI+559R58vO9jjBs+DjecegNmfWYWmt9r1h7P50viO8qdOE2cTnr1X3BO6iqs6ndzGNvvT6HvvO6zB+ApTjq2F52FlU9gGHJnrwyTLiysfALAIlevGQgfEmq2n4Xpc3cbu8Wrk4Fc8Dc234/b5P6cQYTb1P24s7kCc6Z+d9DPa0vX8bypYlnfefUZRDm3STp6jZPbsmR6z73HHaKcBC5CffzdNT/Fk+//COn4LsRSh+LiSX+H+qNHearzfOm4muJ04RLg27mrRrY+EtB2VyYhd6C8lqX5M08JpJ1XjEE516LSyYU+foC3Os9aoYHkvOMrUg1YXKZJNRObMhZmnNYdcx0e7/4GbsRjGC87sVWNxt3peZiRavC08rTQFmG6uvu0+BiMhzMZvU3GYJyH87AV2bKkSWb1xKtxy76L0dqVORZ0AnfdMdcd3IYlq1NV4o1jrsN031+tcNs9UePs3wL+96WshTgoF/UbmYfZRrcSkWuCzcpKXZv3yfWtuHhaHZ5/a7unFd8Vh2xA1djVkEQ7VHctDmyfia3tU8O/NpUYUZqbf5WC+vp61dLSUuzTiIRAlug01gJwfjfSSvCZAz9zHBcA7zeFVLDuOslQ+U10dHxtmSr2xl83Yn+/GRAVUgERQXf64E0Oq+PVmH3sbKx8Z2XOY6vj1Wj8fCOAAC4QhjgBAjS2e3turwxx2pIegxldS3KO2X5/TEkZr3GafexsvLjlxQEv+O//8Qx075nqOC/T+1CNtRBNnBQEUsw4+VCWJi1oNn0D8f2rc7/zZ0w4w/i55x83xa79T7OhAMcFP7Vnqqc6aMstx2BCzNmB25Iegwm3vTvo57WV31ACgPeqLkdMvJVz3fMmE3HMO3M71rb9dMA6zzZOjZ9vRMsHbY5yeutZX3d9bovmTnFeywKs/z0LsD7WXZtaPmjDEx/eBYkd/OxVOoaKeAwp1dN3zLbO+9dlSXyS/rWjjB0e+zzWLnC54sMQp47kEThH3efY1kY3QFFXm3T/ejZCbEPY1nkKyIldQqrwvRnfNZYlL52cxttvxU3d9+UMfnWoStyZuBaNN4c3KKdlGaMgB9/ctitM7TzdMau2X/5gD5BJiFy4xNHZtqpL/WBxbrbcxsmmjNm08/yIUyeq8J2uq7AqnXtDRD/qNt133lSXfrX6Zdyi/tOZ/Jx2u+c9qN0mxADvZclznAp9X4GcZFbjvovx0KenOZ4iqOtSQ9MaTNvzLG6qWNY3mHFnz6VYf8g5nl9P95l9634grkmeCYDaI92182zLmBXToFxI26g1NK0xt0u+tCOQcwiizhtMGbOSH4/jzgV+90gg1wRfaL4/DU+PsWqDuq3zGle9ic6ax/La6AkkOk5DRc16ba5oKCWpRWS9Uqpe+zMmqEtH8wv/gnveW46PY8C4NHDDZ76MWV/4nvY4ANeP1R7fewDo+hT3HFqLjyviGNeTwg272jFt33DMqPxrxwXt0GGVOHTCf3sayXPdaXj/DQDKcW6z9nWi+YqfDfr1TBV+dUU12g+0u4pRTGJIq7TjeE1lDQ6kDjgvJrUn4sVtLYOP6YblVp04P0Zc3T72Sw9fjqeHJx1xSivBgkOPcSQ+/unSzkE3cv2KU76+C/4fn8J+dfACE0/HcfjuOvSM+ADbKwRjexRGbz8VrfI3+KfTnw8kTl7j4UdZ0j3Hri1fxK6OLm2dgDFP5MTJazwAAKlhUNLtuOAnd89D4yl7HZ/9u9v3YeXu1X1xml0zE9df8u+OMnb9zp24YF+H4+XSEMQ0ycWgytKsz8xyDMA0frodVV3tzjhVjEbz7DtcvZ4u6djz6fGorF0PxHIbrqayZBOnZGwkOnr2O+L0uTEzsfnAesd7XvLEPzjiBMAZuzcfhE0SOMw6b9bK76C5Z6cjTgcqa9E4Yqzr5KLba9P+rjgQ3+cqHjZ13p7tpyBRu94Ru7rdR0Ad8qG7dsUL92rjlD/I3Xsj2qc2/QIy6pd9303Vdj4umz7RMXjiSxn7dB+a/3s+7jlk2ME47ekATr0C9+x4ZdCvBzivTV7jAWTKkpJu9wk4l+2Kq3bsw2UdbY7XMw3K2c5S9RSnMZ8DXn3YEaNZX1yM5hHDcx7bMOrreHzdZsf351/PvdIxE/CMCWfgqT+uQLc60Pe+ElKFuZ+d47n9p2vnFUwO/Oll39t5DU1rjINLbttYtrHT1d3HTDvb03fFJgFjYt3OM00u8Rgnm0kZXsuSKemY2jMV3572nDNOY4fr+4Qek866gbbhlUlPZcmPOu+GvQcwa9c2zQs6y5LN5As/6rz3/3iGcfLF98/TfAePPN1TWerZPQ3dw37ruM6LqgykXWFVlkZNsRr4Mg1QeomHzXXlhlNv0L43U1kqdp0HmMtY4+cb8e56l3WFKU5//jXg7WdcJfCD7Eu5bVc8s+NraB4xXDtYM+mzL7qq83TX+Rgq0ZHa43jPghgUnHE6YvgReOYrz7iKXzlggroMNL/wL2h8fzn2xw5uIladVphdPREr92/OOV6RTmcKigz8WNvn+IuqKVhzYNOgZ2wNZhTW8RxKQYC8c0tjdpdg5fDkoF+vKJQCXMTJFI/GkVOAP/3aVUcb8D4r0uax57R9imdHxrA/Fst5H4Cgp997601avbb7v109r02H0Q8xCNK6Zqojdml8AUfgBXzsKGNe4+RHPLyWJdNzxBBHKg1I7OCMLZVOYHgiqb0wB2U4qpBK73eWGwi6Y27ilMbsvZ/ixWHDcpKLZ6Zr8Pwl/+b77AWrOg8xiEo56/Sx9VjZ/qar19MmHXO/wv5S0O55mX+8WhKYPepkrNze4ip2/7xrPxKpDkcSGMNG457Dxxe1zptde6LmfSikEIOKHaxDTHWe7bUpsPipGCCaTo7L61V1WqFxj36Qe6caiTsOPdwxmJUatSwnaRhDHPFYzPc6zzToWIFY5jnUwXqsWhKY/dm5rl8vsGuToSwZO6Ou46Sv887f14lf5g1QDmaWqucyBqWNx/JNT6ILB9ubkhakEXNcg3SzlExMHcag1MSSONDT4a6dt2MnALgaSP5w81HawaXu9mk4ZOxG3+u8k2u+iFd2rHb0CRIxoKff52kqS0EmYGwYJ5f4ECe3kzL8aFeYko66suBHuyL09nhAdZ6uLNlMvvCjzkM6lllVmVePfWbfBHw0/D1Hu0LiFbn1o2VZinSdZ2g/6CZlmBLJl02fiP/ausRTnacbaNENZtqUpajUeSaZvlSno+9uEyddezyomeB+5I/O2ZvGqpFJx3VMRADJLWNB1nkCwWtXvhbIc0cRE9Rl4NwHT8JHcWcPJaYU0i57qKbHWj2HRUXpx2OtnsOUSPR4bjZszsH4HBbxqEkrHIhXOjvasbjrhmRQMYUSQLstgcfn9YFVnCyyQKbY+REn7esFVR4Nn0/YnUMrPsRJ28nRNDCNzxt2nWdVHg1JRwt2rwd9glr3vBZ13iE9aXTFxFVnLew6z494BPYZW9V5Fs9rqvNSaRyQ/DilkULckaxX6QRiFc7VC9rXi3C7wkZQcTa+nus6L40vdSo8fchwT3Ve2NcmHaUE4rIN4sdnbMXielWTSmXL0sGEgWkg2Tww6L7+D+q6Uu7tCl2cbCZlmNi8D1PS0SYZaXM9tuJHGfOjXaGp83RlyXbyRWB1XkB9dD8E1Zeq6UnhQExT5+3dh5WjxrpK7AcxE9z6saa8S7nUeZo4aZPZASblfWmDBFXnmV7PcG6cQX1QTHeQoudjQ6RsqiHTY62ew6Li8+OxVs9h0Ym0ed6ayhpUSyLnWAWARN7gTnVa4ZJ9B1CdTucdT6M2lXsjioLn5vqRwG5BTtITyMxa6Z/0BID9qf3GpGdQMdUvw/fjefW8xgmGwTqbStL0LvyIk/b1giqPhtgF0aCqlgQuGTUV1enc16xIpzWxsytLJsZ3kddA2R8TPNG2wfVKi9DrPJt4WCSndWXJOk7pYOq8PfHcpCcA9MRyZ74Cxanz7MqH9/gn4yMRS8dzjkkaSKSd8Zi7d79FnWdRdxuO747p4pQ7kxwAJNaNWNxdchqIeBlzya+yFLOYYOK+zothxfAKz3VeFMqY7r4PJsqQ9aqJJUO9NunsjuWuRAMycXqiOuaMk6mPbVH/23zGNjMw/fiuxMTQItO18xLjHbEz1nk+TNbSxaknFstJTgOZOq9l19OBtCtM8bCKk+tH2omnEi7LkkJtSn8WflybdHWeriylkcpJegKZ2JlWBgbXHnfPj7Jk05eyaVfY2B031HmHjHCWm1haGycVc5ecBsKPnVW7QpnqvNz/VqIi9L6ULk49sVhOchrItMef+KO7LR+BYuSPgmHTl6pOq8wWJASACeqSMc5QemwCaHqs1XOYGocBPdbqOSyG5t0+b7UksPCIM9G4YyeO6O6BKIUjuntw+7ad+N723GONO3bi5m0fo3FHW97xNizYuSuQhKgfgoqpBBQ7U4XvNU6X7dmrHVy4RHO83GNn6qzZdOK0cUIMl+3tcMbjzV85Y7ejTRM7fVnKJHEsGuauH+lT59nisX7ESUdcvmtTWbKN03eGYJ1nd70yfBYW16ZbJ56N77Xl1mP/d8cOfM8RuzbcsuOT6NZ5FhNXwm9XuFdTWYPqeHXOsWpJeK7zdGUpyDjZDPb78hkHFA/To/M/IpVO4DOfHq3tMC5s2+Xp2mQedPDeJfYjTl4f6/a6Aphj5/raFK/GJZ+9BJWoyD2eVrhsz15nGfvjK47Y2dR5fsVJJ9LtCteP1Nd5SMcg6dxKPZaOY37bPpdlaScW7GzTxuMrEajzTKJQ57luP2TLkte+lE27wpc6z2KWq8182LBjZxOnrl2nOSYiVKQFc/fsy/ncb/5kB+b/7vlQ+1I2Il3nWTw2GRsJlc4tNyodQ0VetG37Uo07dmbuY0AAuMVHyQh/D2r3+1sFtT+u9XPs68TKSjXwUpPevfDy9qCsRiyzN1l1Ze5NebpjQGebMyg2JI7mYVU5+zSd0dGBlSNH5C1hsotpdWIY2lPOO8/q2Nx0xK+Y5t+MqCKtIGLY8zip2QJDsydo45Gz9DdP8iFOzcOHOfc8S1ejOZHOeb0zOvdj5YhhrmLnR5wC24Pa9Ni27Vg5crh2ed3ymkNz9v+MQxCTmLs47dqNWfucsxqsVn5qytINu9qByhFoPKTK1T7Gn+s+DK8ktuWO/BuWtXm90VKgdZ4mTrr3HEvHEes4HRUjfxvBOi+N2Z09WJlMuIpdRXwYPlXuZmEUo85zPFZbd6dx0d59+PnIWqT7zfyJpeOoH3M+Xtu12vc4mVZtauu8jgOerldRrvMGtR+/22vTjO85b5RkuimXDUOd5zVOQdV5xbi3ha59NGr3Ufi4ptVRxsbl3eD48LZ6LFWv4PnYbs0+z+5n9xuvTXDuS7wP1Vg8Zpir92EqS7Zxym+P+bHPu24P6nhaIaZp5523twv/b2Slo/4/Yt8x+OiQLbl1niTwFxWT8WbHazk36vqLqsnY8tbt+MGoEX2f5/Vt7bigw32c3NZ5N+xqR3d8GP7vodWe4qQTiXaFad9cU58wFkN3v0Suqc6bJ5/FodvX5cTof7d9ijkde6yShro4zdy3H6uHVxetzgOK0K7Q3tvC0M8fW4+Vba85ytLsUSc7byI5aor2ZsFBtSuMbfQA6jxAf3PhwNp5FmXJNk7rXorh5MoH8Z/9ytN1be24UFPn+dGXKuc6L4Y4VDqds5Ivlo7jf1X9GX5z4HVXdR7a5mFS5+vYOfbVvmtT/Y7xOD2+CT84dLjHvpRob/ZerrgHdZlw3I3YdJdi011xDY81PofLOwT7dSdwz3dyffhyNA9PDtw52NOBWZ+9GM1/fNJV8sxKchTQ06m/o+3vHnEcb558Ju7Z8/qgY4ojT3d9Q73GzzcGFifXj33/DQBK2wlsPqQ2Nx7tewER3FMzwv/kmS5OsUSmpZXqyo3dhUsy/37utpw7Eje3vW5VlrzGKdTYrfwOmnt2OuL0xa4Enk6onI7H37ftQSwmuLd25OAbudA3qtLIm4FQoCzhwiXamLy7fZ/jbtSPv3sB6tSPcxoYFZ8eja01HwEx942fF7e8WNw6zxCntALuHXXw2DfbPsWne07D6FEt/td5QDZ6/doSpjjFEpk6Or9Mf3Gx69jtGj0VT3x4l6sb9UaizvtkK9DhjFHDvjjWDJOcTsc32z6F6vw8qke+7Huc2jEStYked3WeH9erCNd5Nnd317YVTNemz17s/NysuSxLBY4/OP7z+FlqU065uf6Sf3fE6YzD3N/AybbOC6qMHfbGFkeC8rq2dsTEmQTWlbFr2/bgSx37UdlvkLUnXo0K6xtmu6/zdGVs3ZTv4q63fo1PRrXkJMpPnzQa/7X3mQHLkm2cunefgn965ifaG4mtbfupp9h9d81P8eT7P0I6vgux1KH4xiH1OGbzk46y9Be7e/CbEWlHnE7fG8PLtRU5j7+ufS9m7u1EpRyMU6eqxH6pwqHY6zJEcUDplrG7L2ON6hr8Pv5eTlth9PZTMaKqAltq1w26zotKu+Ld9c85rrHXb3lG264ABPccWuOizsv7fAeMh9OBRA3SXZ1IysFy06kqsXH0LJyys9lx/NGj/xKPqT8OeG2yqfMi1ZfKv8buNdxEUDOh5ob2vZjVecB5na9Ieu5LtWMkKtUBDOsXjwMqDhHJqWMLtdEBd31b25sZfm/Gd7WfZ2Cxyx+c/syXMWvDcm1ZMsXpvH37Ee+XtO6JVyMVq0ZVd7ureJi3j7ZoV1y4BEs2veyoF44ZO9zx/lrSn3XdHo9CnWe68elLXQ9h/fB9ruq8nW31+Er8xZzvfFoBMdcjA+Y+b0fyCAz7zlvun6jEMUFNQ8NdJwG7N7t7rEVDqcCTwFHhG5KZOPlS4LVl+uMaKza0YvHqTdja3onxtUnMnzkZc6bWaR9rU4kHxfX5mmLkSzxMLOKkO2aIka0oxMm115YBv7je0XBp765ArdvOoYWd6RFISlfOBb9DVeLnqTNwxehNnsqSzqQFzdoFnYlDNmDSZ90nYIpOEydHUj+rBzFUeN1prdAA3NvPuIsT4Cl2DU1r8En6144G5rDKOHDoL/uSJBdP+jvcetbXvb1fPxjKUltXHKPkU8fDe1QMFS73izUO7OQ1ljtVJd6YdjumH32o+3h4LGNAidV5gP49P3dbMO0KP8qS5vi6Y67DFeuOQmf3wfNIJuJYNHeK9ppsk8yKSpzWrfohJr66GIepHdgmY3A4tuvvCwi72WTa2CVHoaerIyeB3ROvRsXUyz3XeV7beTZxamhag9Z258BJXW0Saxec5fZTck/zOaSfulq7tZFNnEwJmPzjfTEyDey4jN2kR4Zr2woC4P0m5+fsR1IlLCs2tGLhU6876orfx+e537/dtu2eSLoawGlU16Ctows3VSzDeNmJrWo07uy5FOsPOQd3n/B2TvnffOp8tE68IJCyBET0etVYC5v7/PhB1674xVEL8NsP2nAjHuuL092Yh8vqj8T0d+/1vf0wFONkUz9qJyIMpl3hkqk9fuiwShw64b8jV+eZ+n7vVX3NdYLZj76UbmCnQ1XizsS1aLz5u56eu5QwQU1Dgy4R4AfbjqRHpkajqYNZbFbna0jW+BIzyzjZdA6HJF0H88m/sxol1nMOGBTqjNh0nt3GNPTOepDy4qR2b/aerDGVpUIDcCExNTDzRarODChZY+qMvDt+NoZ/+FxOB376RdewzhsE1VhrdbM9VwIsS2VVt7llMzmhkPy2SHam8+Mtf3IkYKpPnYfn39peMmXJVG+aEq1B6LjjeAzr/Mhx3I8E9c70CHSiOidGM758LebE13qaGLJ49aayLU+muuLl6hswDtv9f8Gaif0G/Q7GY90HuxwJ50t/PcH197XU+ky+MNR5Vu08E037rxNVWNbzlzg7ttHRRu8tJ4OtC03tkii3VzxPyrJgOxEBgKM8Tb/oGm/vwyAK1xUbftR5fvSlbtj3N1CAo8/7i/SMSH5uQWGCmoaO/ERA1z79EiabpX8hJ2VKrYNpfb5eZ6n5kDwbkg1aH3zceKz7jovFgMGKVIPneNjEtJzin9/AfFau1SYBolzn2TDVNzpRrTMBc7ImLTHENPv05XdGOlQlFnZ/A/fMm8o6L0BWdV6BWbioHB5KWSq1DqMvTAPfNkvYDcmzhqfHaOub/E0Mol6WItGufG0ZelZe55yNXjnMdZx2YQSqVVfOFg8dqhILur+BVekZOY+1eW+m+vHiaXV4cn1rWdabprpiduwl/FvywZw4GZMylu2KFamGnPbKmceP1X6+1YkYdnV0O55VF9NIfLfD5scqR4u+1A3bL8TKvPIFeL+ulGK582NS1gFUud62Y2d6BEZXpVxNRGideEGo/aBSK3um9/zw9A8x/fVbPW5V5L4vZWpXRPVzC0qhBHWF7mDeLz8I4AIA25RSJ2WPXQKgEcCfAThNKdWSPX40gD8A2JT99ZeVUt/M/mwagIcAJAE8DeAGpZQSkVEAHgdwNIAPAFyqlNo1iPdJlOn09e/4mTouhfZ0NM2KDik5s9WQfDEdLzbr882PUa/8OMUSSCk49uOqOP+OzH9MSRkXcVq8elPOBQoAOrtTWLx6U9EbP1G2+dT5qFl/c07nsEtlbrTUP05IJIGB4tTPnOzfXkbybWLa+/+ozs5wK7+x1dreiVsqL0ZTYmnunqoRr/NszJ852dHANIlqnQkAw86/TZ+s0SxL70QVlqX0s5dw8lms83yim030fNclWJRY6txjE5KzP27BMnb+HaGVpfG1SW3HZ3xtMpTXL4rez1a3tYamXaHdd7233suL09ZHmrUvmd9tjXpZ0tWbyUQc82dODu8kTr400+nsF6cKQ5xSUoFUGo49qN+Zlnl8/xmC/9p9iSM5DdjV/6b68fm3tmPR3Cmu2wpRnvWZz1RX/KrqTCzojuWsGvgVpuKyxP9o2xU9G37maguc/IkIre2d+NnLf9KWpaqKGJKJuKvva6n1mXxhqPNufWyD++uVRV+qpWkNEMB1xVTuHn1lM1J5EyijUsdataUMcVq06k3cpO4bME4dqhL3Vn4DjRee6HiOY/rFZ1z2z41Na1yfmx9twkhcVyyY+n7Tp54H5G+Bd9y5Vvf+cAwYTJmP6Ya+1PyUPlEe1c+tGAZMUCOTVP4+gIf7HXsDwFwAP9Q8/l2l1Cma4z8A8HcAXkEmQX0egF8CWADgOaVUk4gsyP7/Oy7Pn6gwU8fl5EuBI0+P5AzBUutg+nK+mjitO+Y659La9DzMSDVkLjIeYjUkG7Q+mH7RNViHvOVj0+br97a1THLOmVrnqeFpG1OvrxcFugbmz7s+jxGVFWisebJk6jwbugZmR1ePdrZVVOtMAOZkjSZObxxzHZrWHYVbu/Jmulg0ZlnnFaYb7Fn41Ouorj4TC/Y7l2KOGlaJxuHRK2Ol1mH0jWngG/C0D76pfaMT5bIUmUFZl3GKn30LXs3f+mFav6Xq2b/HAVjvQ/KsUP3otq1gqkMARLKtYaorRDLtiJ/j8zmPf7vyREe7YkWqAS/9Nulsp9ddizkX/HvO7y/WJM9Ma7h3d3bjrstOcfV9LbU+02DoBz6cZanl6TFYsMfiegW4uj4FdV0xlbv85PRAjw+TH5OyfvJIM9pizm0NgdzY3Y15mDHrauBkd31em3Pzo00YmeuKBWN9rrs2mdpzeccP3vvjor5fTa6LY9HEVu1rleLnFjZXW3xkZ0b/V+8M6n7HXwDwj3kzqHWPOwLA80qp47P//yqALyilrhGRTdl/f5R93AtKqQFrPG7xQeWq1JZiB3W+QS4dKrVlSTSwoRhTP5bz+zHjq9izxkqtzhwMr5/xUCwfNkyfT20ygQM96ZL6bhW7PJYTXd1iWPiLun77Fg/1zz7M72AxlqqXy37Vuvfx7cc3um5X2Hxubu8fYfr9Qu+hnK//Udm+Logybfr+xEW0SeoolCU/2lKF2hvDqypCaeexTegffpaD42mLj0GYJCIbAOwBcLNS6n8A1AHY0u8xW7LHAOBwpVTvRowfAzg8gHMiKhmlNrIW1PkGOeNvyM4yK2NDMaZeZw75MeMrCrPGSq3OHAyvM/6HYvmwYbqu2MzksxVUEq8cVodEha5uMe2be+bxY0OvC4P6Dnl53sFcE7y8nh/1v039aHp/pm2nojDr00RXV5gS7bp2hU073dRe0e3nbnNdKvfrf1S2rwviumIqd6Y9qKPQXjGd85nHj0VD0xpXn7vpORovOjG0dh7bhP7hCkX/+Z2g/gjAkUqpndk9p1eIyIlufzm7J7VxgFVErgZwNQAceeSRnk+WKKpKrYMZxPkGuWyv3Bu0Q9FQjKnXBqYfe9BFZW/jUqszw2YqHwBcd6rKWaHrje675TUxGIWBHXJHF//6o0ZpZ9CGWRcG9R3y+ry2n4Mf78Nr/W/TfjC9P9Osz1LbasKmXWHTTi+UjHz+re2erkHlfP0vxvZ1Ya2AKFTudHVsFGLsZtByoDosqP6KzfMOxT5TUIbCNkNh83WLD83vvQDgHwG0glt8EJGFcl+2R+QHLx0JP7YI8eM5qDhYxx7k1zJqwF2Hj0tCy0/YdWFQ3yGvz2v7OZRaWSi0VYXuxn6lWJ+6bVfYXkO4/ZC9sMsH2wX2Sq0OI3+xzAxOaFt8iMhYAG1KqZSIfAbAcQDeU0q1icgeETkdmZskXgHg3uyvrQJwJYCm7N8r/TwncoeNBooaju4SDczLbBk/Rv05c6B0+TXjsxzaD37MoGxc9WbOftWFZlFxSWj5CbsuDOo75PV5bT+HUisLpvdXTnuQu21X2LbTy3mmc1DC3oohKqviSkmp1WGDUQ7tvKAwX+G/ARPUIvIogC8AGCMiWwDcCqANmQTzWADNIrJRKTUTwBkAbhORbgBpAN9USrVln+paAA8BSAL4ZfYPkElMLxORqwB8CCC8250TAC41pegqxcYsL+JUKvzo+HAfu9LlR6eqnNoPbq83ps+nvbPbcaw3cZ1/TeDATvkJuy4M6jvk9XltP4dSKwuF3l8ptlm9GorvOUxhb881FJKtfiu1OsxWObXzgsJ60F+xgR6glPqqUuoIpVRCKTVBKfWAUmp59t9VSqnDs8lpKKWeVEqdqJQ6RSl1qlLqF/2ep0UpdZJS6hil1LdUdm8RpdROpdTZSqnjlFJf7JfQppAUGi0lIvd6L+Kt7Z1QOHgRX7GhtdinRuQwZ2odFs2dgrraJASZGWC2S9L8eA4qDlPnyaZTNRTbD7adzvbObsc14czjxyKZiOc8jgM7pc1UFwKZRNKkBc1oaFrjW3tg/szJgXyHvD6v7TUhqPcRFF7zKGxzptZh7YKz8H7TrL4tI2z6Gis2tLqug/xoFww1pVaH2RqK7TwqLld7UEcR96D2D/cQpSgoh5nH3IeMiLwKqy70Y9+8odh+MH1u1YkYdnU4Z1HrlNN2AGQW9N6UQdUVYbfHyqH9RxQWm77GYPYJ53669sq5DhuK7TwKXmh7UFNpKvelKRR95bJ8iEvjiMiLMOtCP/bNG4rth0JLrvM79iZb2zu5JDRigkgwBL2fa1DfobC/mywLRO7Z9DVs6yDupzs45VyHDcV2HhUXE9TEPUSp6Mrlphy8iBORF2HXhV47VUO1/VDoc+vfse/o6tHOquY1IVqCGhjioDUR+c2mrzGYOijKydZynqkcVUO1nUfFM+Ae1FT+uJ8aFVu5dOLKfR8yIgpWqdWFbD/kyt8r9NYLT+Q1oQQEtccm93MlIr/Z9DWKUQfZ7Hlt+7y8z0/42M6jsHEGNQGI9mgplb9ymXnMpXFU7jh7JVilWBey/WDGa0JpCGpgiDPPiMhvNteVsOugILcpK8XVtuXSZmY7j8LEBDURFV05deJ4EadyVS57xUdZOdWFlMFrQvQFNTDEAQoiCoLb60rYdVCQSeRSW2HGNjPR4DBBTURFx04cUfSV4uyVUsO6kCh8QQ4McYCCiIopzDooyCRyqa0wY5uZaHCYoCaiSGAnjijaSm32SqliXUhBKJelxkHgwBARkXdBJpFLbYUZ28xEg8MENREREQ2o1GavEFEGlxoPLAoDQ6ZBhCgMLkThHIgo2oJejQKUzkAi28xEgyNKqWKfw6DU19erlpaWYp8GERHRkJCf5AIyHQ/ezZso2hqa1mg7ynW1SaxdcFYRzojymerXi6fV4cn1rUWtd1n3E5FbHMzKYL1JZCYi65VS9bqfcQY1ERERDajUZq8QUQaXGkefab/SR1/ZjFTeZKKw9zHlXqpE5FYUVqNEAdvMRIPDBDURERG5wo4HUenhUuPoMw0W5CenB3p8EDjAQURkj21mInuxYp8AEREREREFY/7MyUgm4jnHonxzqaHINFgQF7F6fBBMr8UBDiIiIvITE9RERERERGVqztQ6LJo7BXW1SQgye09zH8xoMQ0ifPVzE4s+uMABDiIiIgrDgFt8iMiDAC4AsE0pdVL22CUAGgH8GYDTlFIt2ePnAGgCUAmgC8B8pdSa7M9eAHAEgN71YOcqpbaJSBWAhwFMA7ATwGVKqQ98en9EREREREMalxpHW6H9SuuPGlXUfUy5lyoRERGFQZRhb7O+B4icAeBTAA/3S1D/GYA0gB8C+Md+CeqpAD5RSm0VkZMArFZK1WV/9kL/x/Z7/msBnKyU+qaIzAPwZaXUZQOdeH19vWppaRnoYURERERERERERERURCKyXilVr/vZgFt8KKVeBNCWd+wPSqlNmsduUEptzf73TQDJ7AzpQmYD+En23z8HcLaIYcM1IiIiIiIiIiIiIiobQe5BfTGAV5VSB/od+7GIbBSRf+mXhK4DsBkAlFI9AHYDGB3geRERERERERERERFRBASSoBaREwHcAeCafocvV0pNAfCX2T9fH8TzXi0iLSLSsn37dn9OloiIiIiIiIiIiIiKwvcEtYhMALAcwBVKqXd7jyulWrN/7wXwCIDTsj9qBTAx+7sVAGqQuVmig1LqfqVUvVKqfuzYsX6fOhERERERERERERGFyNcEtYjUAmgGsEAptbbf8QoRGZP9dwLABQDeyP54FYArs//+CoA1aqA7NxIRERERERERERFRyRswQS0ijwL4DYDJIrJFRK4SkS+LyBYAfwGgWURWZx/+LQDHArglu9f0RhE5DEAVgNUi8hqAjcjMmv5R9nceADBaRN4B8A8AFvj4/oiIiIiIiIiIiIgooqRUJyvX19erlpaWYp8GERERERERERERERUgIuuVUvW6nwVyk0QiIiIiIiIiIiIiooEwQU1ERERERERERERERVFR7BMgIiKiaFmxoRWLV2/C1vZOjK9NYv7MyZgzta7Yp0VERERERERliAlqIiIi6rNiQysWPvU6OrtTAIDW9k4sfOp1AGCSmoiIiIiIiHzHLT6IiIioz+LVm/qS0706u1NYvHpTkc6IiIiIiIiIyhkT1ERERNRna3un1XEiIiIiIiIiL5igJiIioj7ja5NWx4mIiIiIiIi8YIKaiIiI+syfORnJRDznWDIRx/yZk4t0RkRERERERFTOeJNEIiIi6tN7I8TFqzdha3snxtcmMX/mZN4gkYiIiIiIiALBBDURERHlmDO1jglpIiIiIiIiCoUopYp9DoMiItsBfFjs8yiSMQB2FPskiKissF4hoiCwbiEiv7FeISK/sV4hCsdRSqmxuh+UbIJ6KBORFqVUfbHPg4jKB+sVIgoC6xYi8hvrFSLyG+sVouLjTRKJiIiIiIiIiIiIqCiYoCYiIiIiIiIiIiKiomCCujTdX+wTIKKyw3qFiILAuoWI/MZ6hYj8xnqFqMi4BzURERERERERERERFQVnUBMRERERERERERFRUTBBXWJE5DwR2SQi74jIgmKfDxGVJhH5QEReF5GNItKSPTZKRJ4Vkbezfx9a7PMkougSkQdFZJuIvNHvmLYekYwl2fbLayJyavHOnIiiylCvNIpIa7bNslFEvtTvZwuz9comEZlZnLMmoigTkYki8ryI/F5E3hSRG7LH2WYhihAmqEuIiMQB/AeA8wGcAOCrInJCcc+KiErYmUqpU5RS9dn/LwDwnFLqOADPZf9PRGTyEIDz8o6Z6pHzARyX/XM1gB+EdI5EVFoegrNeAYC7sm2WU5RSTwNAth80D8CJ2d+5L9tfIiLqrwfA/1FKnQDgdAB/n60/2GYhihAmqEvLaQDeUUq9p5TqAvAYgNlFPiciKh+zAfwk+++fAJhTvFMhoqhTSr0IoC3vsKkemQ3gYZXxMoBaETkilBMlopJhqFdMZgN4TCl1QCn1PoB3kOkvERH1UUp9pJR6NfvvvQD+AKAObLMQRQoT1KWlDsDmfv/fkj1GRGRLAXhGRNaLyNXZY4crpT7K/vtjAIcX59SIqISZ6hG2YYjIi29ll9o/2G8LMtYrRGRFRI4GMBXAK2CbhShSmKAmIhqaZiilTkVmCdvfi8gZ/X+olFLIJLGJiAaF9QgR+eQHAI4BcAqAjwD8W1HPhohKkoiMAPAkgBuVUnv6/4xtFqLiY4K6tLQCmNjv/xOyx4iIrCilWrN/bwOwHJklsZ/0Ll/L/r2teGdIRCXKVI+wDUNEg6KU+kQplVJKpQH8CAe38WC9QkSuiEgCmeT0z5RST2UPs81CFCFMUJeWdQCOE5FJIlKJzE1BVhX5nIioxIjIcBEZ2ftvAOcCeAOZ+uTK7MOuBLCyOGdIRCXMVI+sAnCFZJwOYHe/ZbVEREZ5e79+GZk2C5CpV+aJSJWITELmhma/Dfv8iCjaREQAPADgD0qpf+/3I7ZZiCKkotgnQO4ppXpE5FsAVgOIA3hQKfVmkU+LiErP4QCWZ9pqqADwiFLq/4nIOgDLROQqAB8CuLSI50hEEScijwL4AoAxIrIFwK0AmqCvR54G8CVkbmLWAeBvQj9hIoo8Q73yBRE5BZnl9x8AuAYAlFJvisgyAL8H0APg75VSqSKcNhFFWwOArwN4XUQ2Zo/9E9hmIYoUyWy1Q0REREREREREREQULm7xQURERERERERERERFwQQ1ERERERERERERERUFE9REREREREREREREVBRMUBMRERERERERERFRUTBBTURERERERERERERFwQQ1ERERERERERERERUFE9REREREREREREREVBRMUBMRERERERERERFRUfz/ZSzOpViYbo8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1800x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(25,2))\n",
    "\n",
    "plt.scatter(np.arange(len(y)),\n",
    "            y)\n",
    "plt.scatter(np.arange(len(y)),\n",
    "            (np.asarray(language)*betas[1] + betas[0]))\n",
    "plt.scatter(np.arange(len(y)),\n",
    "            (np.asarray(string)*betas[2] + betas[0]))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(25,2))\n",
    "\n",
    "plt.scatter(np.arange(len(y)),\n",
    "            y)\n",
    "plt.scatter(np.arange(len(y)),\n",
    "            (np.asarray(language)*betas[1] + np.asarray(string)*betas[2] + betas[0]))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Subtraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "208.873195130127"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Language - string\n",
    "cope = betas[1] - betas[2]\n",
    "cope"
   ]
  }
 ],
 "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": "3.9.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}