Simulation Fishers.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "06507b89",
   "metadata": {},
   "source": [
    "#### Distance Method\n",
    "\n",
    "- Bigger beta or smaller sigma squared\n",
    "- Beta be fixed\n",
    "- number of interactions are fixed\n",
    "\n",
    "#### Score method\n",
    "\n",
    "- Same parameters as the distance method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "21b522c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sample import simple_sample_generator\n",
    "from estimators import ClusteredSegmentation\n",
    "from BarcodeScanner import cluster_barcode_scanner\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.metrics import auc, RocCurveDisplay, confusion_matrix, ConfusionMatrixDisplay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "18458cda",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "from sklearn.metrics import confusion_matrix, f1_score, roc_auc_score, RocCurveDisplay\n",
    "def simulation(sample_size, num_variables, beta_range = None):\n",
    "    num_all_interactions = 2\n",
    "    k = random.randint(1, num_all_interactions - 1)\n",
    "    if beta_range:\n",
    "        rng = sample_generator(num_variables, sample_size, k, beta_range = beta_range)\n",
    "    else:\n",
    "        rng = sample_generator(num_variables, sample_size, k)\n",
    "    cluster = ClusteredSegmentation(n_clusters = 2**num_variables)\n",
    "    cluster.fit(rng.barcode.reshape(-1,1), rng.y)\n",
    "    scanner = cluster_barcode_scanner(rng, cluster, False) # <- need to change to False later\n",
    "    scanner.set_beta_sum_range(1, 2)\n",
    "#     ranking_table = scanner.get_distance_ranking(normalize = False)\n",
    "    ranking_table = scanner.cdf_ranking\n",
    "    true_parameters = [x.replace('X','beta') for x in rng.interactions.keys()]\n",
    "    actual = ranking_table.coefficients.isin(true_parameters)\n",
    "    predicted = ranking_table.ranking.isin(list(range(1, k+1)))\n",
    "    return (actual.to_numpy(), predicted.to_numpy())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ddcf9fa",
   "metadata": {},
   "source": [
    "### Simulation 1: \n",
    "\n",
    "* Sample size : 5,000\n",
    "* beta: Unif(3, 8) * Bern(.5) * (-1)\n",
    "* Number of binary variables: 3\n",
    "* Number of intercation terms: between 1 and 4\n",
    "* $\\sigma^2$ : 5\n",
    "* Number of iterations: 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1e2ca878",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ True, False, False, False]), array([ True, False, False, False]))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sample result\n",
    "simulation(5000, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a31d8ad5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 10 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:    6.3s\n",
      "[Parallel(n_jobs=-1)]: Done  12 tasks      | elapsed:    8.4s\n",
      "[Parallel(n_jobs=-1)]: Done  21 tasks      | elapsed:   11.7s\n",
      "[Parallel(n_jobs=-1)]: Done  30 tasks      | elapsed:   14.3s\n",
      "[Parallel(n_jobs=-1)]: Done  41 tasks      | elapsed:   18.9s\n",
      "[Parallel(n_jobs=-1)]: Done  52 tasks      | elapsed:   23.6s\n",
      "[Parallel(n_jobs=-1)]: Done  65 tasks      | elapsed:   27.3s\n",
      "[Parallel(n_jobs=-1)]: Done  78 tasks      | elapsed:   31.5s\n",
      "[Parallel(n_jobs=-1)]: Done  93 tasks      | elapsed:   37.7s\n",
      "[Parallel(n_jobs=-1)]: Done 108 tasks      | elapsed:   42.6s\n",
      "[Parallel(n_jobs=-1)]: Done 125 tasks      | elapsed:   49.3s\n",
      "[Parallel(n_jobs=-1)]: Done 142 tasks      | elapsed:   53.9s\n",
      "[Parallel(n_jobs=-1)]: Done 161 tasks      | elapsed:  1.0min\n",
      "[Parallel(n_jobs=-1)]: Done 180 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=-1)]: Done 200 out of 200 | elapsed:  1.2min finished\n"
     ]
    }
   ],
   "source": [
    "from joblib import Parallel, delayed\n",
    "\n",
    "sample_size = 5000\n",
    "num_variables = 3\n",
    "\n",
    "result = Parallel(n_jobs = -1, verbose = 10)(delayed(simulation)(\n",
    "    sample_size = sample_size, num_variables = num_variables) for _ in range(200))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d69d68e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_true_interactions = [np.mean(r[0]) for r in result]\n",
    "prediction = np.concatenate([r[1] for r in result])\n",
    "actual = np.concatenate([r[0] for r in result])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c983b0bc",
   "metadata": {},
   "source": [
    "#### ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f776d307",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x7fd500cec5e0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "RocCurveDisplay.from_predictions(actual, prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ddefd23e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7fd501b84eb0>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = ConfusionMatrixDisplay(confusion_matrix(actual, prediction))\n",
    "ds.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b120e560",
   "metadata": {},
   "source": [
    "### Simulation 2: \n",
    "\n",
    "* Sample size : 5,000\n",
    "* beta: Unif(3, 8) * Bern(.5) * (-1)\n",
    "* Number of binary variables: 4\n",
    "* Number of intercation terms: between 1 and 10\n",
    "* $\\sigma^2$ : 5\n",
    "* Number of iterations: 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "88940037",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 10 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:   11.2s\n",
      "[Parallel(n_jobs=-1)]: Done  12 tasks      | elapsed:   16.1s\n",
      "[Parallel(n_jobs=-1)]: Done  21 tasks      | elapsed:   25.2s\n",
      "[Parallel(n_jobs=-1)]: Done  30 tasks      | elapsed:   31.4s\n",
      "[Parallel(n_jobs=-1)]: Done  41 tasks      | elapsed:   43.0s\n",
      "[Parallel(n_jobs=-1)]: Done  52 tasks      | elapsed:   49.1s\n",
      "[Parallel(n_jobs=-1)]: Done  65 tasks      | elapsed:   59.2s\n",
      "[Parallel(n_jobs=-1)]: Done  78 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=-1)]: Done  93 tasks      | elapsed:  1.4min\n",
      "[Parallel(n_jobs=-1)]: Done 108 tasks      | elapsed:  1.6min\n",
      "[Parallel(n_jobs=-1)]: Done 125 tasks      | elapsed:  1.8min\n",
      "[Parallel(n_jobs=-1)]: Done 142 tasks      | elapsed:  2.0min\n",
      "[Parallel(n_jobs=-1)]: Done 161 tasks      | elapsed:  2.3min\n",
      "[Parallel(n_jobs=-1)]: Done 180 tasks      | elapsed:  2.5min\n",
      "[Parallel(n_jobs=-1)]: Done 200 out of 200 | elapsed:  2.8min finished\n"
     ]
    }
   ],
   "source": [
    "sample_size = 5000\n",
    "num_variables = 4\n",
    "\n",
    "result = Parallel(n_jobs = -1, verbose = 10)(delayed(simulation)(\n",
    "    sample_size = sample_size, num_variables = num_variables) for _ in range(200))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4ddd7e6",
   "metadata": {},
   "source": [
    "### ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "62d8ea49",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x7fd501b7e220>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction = np.concatenate([r[1] for r in result])\n",
    "actual = np.concatenate([r[0] for r in result])\n",
    "RocCurveDisplay.from_predictions(actual, prediction)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "138fa4b7",
   "metadata": {},
   "source": [
    "### Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "855f3a83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7fd5200d4b50>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = ConfusionMatrixDisplay(confusion_matrix(actual, prediction))\n",
    "ds.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "40803c47",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Count'>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "num_true_interactions = [np.sum(r1) for r1, r2 in result]\n",
    "sns.histplot(num_true_interactions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11159c12",
   "metadata": {},
   "source": [
    "## Simulation 3"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02a6a09c",
   "metadata": {},
   "source": [
    "* Sample size : 15,000\n",
    "* beta: Unif(3, 8) * Bern(.5) * (-1)\n",
    "* Number of binary variables: 5\n",
    "* Number of intercation terms: between 1 and 25\n",
    "* $\\sigma^2$ : 5\n",
    "* Number of iterations: 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "37f75e48",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 10 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done  12 tasks      | elapsed:  2.3min\n",
      "[Parallel(n_jobs=-1)]: Done  21 tasks      | elapsed:  3.7min\n",
      "[Parallel(n_jobs=-1)]: Done  30 tasks      | elapsed:  5.1min\n",
      "[Parallel(n_jobs=-1)]: Done  41 tasks      | elapsed:  6.6min\n",
      "[Parallel(n_jobs=-1)]: Done  52 tasks      | elapsed:  8.3min\n",
      "[Parallel(n_jobs=-1)]: Done  65 tasks      | elapsed: 10.1min\n",
      "[Parallel(n_jobs=-1)]: Done  78 tasks      | elapsed: 12.2min\n",
      "[Parallel(n_jobs=-1)]: Done  93 tasks      | elapsed: 13.8min\n",
      "[Parallel(n_jobs=-1)]: Done 108 tasks      | elapsed: 15.9min\n",
      "[Parallel(n_jobs=-1)]: Done 125 tasks      | elapsed: 18.2min\n",
      "[Parallel(n_jobs=-1)]: Done 142 tasks      | elapsed: 20.6min\n",
      "[Parallel(n_jobs=-1)]: Done 161 tasks      | elapsed: 22.8min\n",
      "[Parallel(n_jobs=-1)]: Done 180 tasks      | elapsed: 26.0min\n",
      "[Parallel(n_jobs=-1)]: Done 200 out of 200 | elapsed: 28.7min finished\n"
     ]
    }
   ],
   "source": [
    "sample_size = 15_000\n",
    "num_variables = 5\n",
    "\n",
    "result = Parallel(n_jobs = -1, verbose = 10)(delayed(simulation)(\n",
    "    sample_size = sample_size, num_variables = num_variables) for _ in range(200))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "81f78bfd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x7fd510c96190>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction = np.concatenate([r[1] for r in result])\n",
    "actual = np.concatenate([r[0] for r in result])\n",
    "RocCurveDisplay.from_predictions(actual, prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "8c47d59d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7fd510c96a60>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = ConfusionMatrixDisplay(confusion_matrix(actual, prediction))\n",
    "ds.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "c1960617",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Count'>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "num_true_interactions = [np.sum(r1) for r1, r2 in result]\n",
    "sns.histplot(num_true_interactions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6dcb87c",
   "metadata": {},
   "source": [
    "## Simulation 4"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1994a927",
   "metadata": {},
   "source": [
    "* Sample size : 15,000\n",
    "* beta: Unif(1, 2) * Bern(.5) * (-1)\n",
    "* Number of binary variables: 5\n",
    "* Number of intercation terms: between 1 and 25\n",
    "* $\\sigma^2$ : 5\n",
    "* Number of iterations: 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "c8523cd9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 10 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done  15 out of  30 | elapsed:  3.2min remaining:  3.2min\n",
      "[Parallel(n_jobs=-1)]: Done  19 out of  30 | elapsed:  3.9min remaining:  2.3min\n",
      "[Parallel(n_jobs=-1)]: Done  23 out of  30 | elapsed:  4.6min remaining:  1.4min\n",
      "[Parallel(n_jobs=-1)]: Done  27 out of  30 | elapsed:  4.9min remaining:   32.5s\n",
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:  5.2min finished\n"
     ]
    }
   ],
   "source": [
    "sample_size = 15_000\n",
    "num_variables = 5\n",
    "beta_range = (1, 2)\n",
    "result = Parallel(n_jobs = -1, verbose = 10)(delayed(simulation)(\n",
    "    sample_size = sample_size, num_variables = num_variables, beta_range = beta_range) for _ in range(30))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "8b43aeba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x7fd5302245e0>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction = np.concatenate([r[1] for r in result])\n",
    "actual = np.concatenate([r[0] for r in result])\n",
    "RocCurveDisplay.from_predictions(actual, prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "bd159546",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7fd5430dd4c0>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = ConfusionMatrixDisplay(confusion_matrix(actual, prediction))\n",
    "ds.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1a9dad5",
   "metadata": {},
   "source": [
    "#### Adding hypothesis tests including three betas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "9fd81fa1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def simulation(sample_size, num_variables, beta_range = None):\n",
    "    num_all_interactions = 2**num_variables - num_variables -1\n",
    "    k = random.randint(1, num_all_interactions - 1)\n",
    "    if beta_range:\n",
    "        rng = sample_generator(num_variables, sample_size, k, beta_range = beta_range)\n",
    "    else:\n",
    "        rng = sample_generator(num_variables, sample_size, k)\n",
    "    cluster = ClusteredSegmentation(n_clusters = 2**num_variables)\n",
    "    cluster.fit(rng.barcode.reshape(-1,1), rng.y)\n",
    "    scanner = cluster_barcode_scanner(rng, cluster, False) # <- need to change to False later\n",
    "    scanner.set_beta_sum_range(1, 3)\n",
    "#     ranking_table = scanner.get_distance_ranking(normalize = False)\n",
    "    ranking_table = scanner.cdf_ranking\n",
    "    true_parameters = [x.replace('X','beta') for x in rng.interactions.keys()]\n",
    "    actual = ranking_table.coefficients.isin(true_parameters)\n",
    "    predicted = ranking_table.ranking.isin(list(range(1, k+1)))\n",
    "    return (actual.to_numpy(), predicted.to_numpy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "9834d79b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 10 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:   59.4s\n",
      "[Parallel(n_jobs=-1)]: Done  15 out of  30 | elapsed:  2.3min remaining:  2.3min\n",
      "[Parallel(n_jobs=-1)]: Done  19 out of  30 | elapsed:  2.9min remaining:  1.7min\n",
      "[Parallel(n_jobs=-1)]: Done  23 out of  30 | elapsed:  3.4min remaining:  1.0min\n",
      "[Parallel(n_jobs=-1)]: Done  27 out of  30 | elapsed:  4.2min remaining:   28.0s\n",
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:  4.6min finished\n"
     ]
    }
   ],
   "source": [
    "sample_size = 15_000\n",
    "num_variables = 5\n",
    "beta_range = (1, 2)\n",
    "result = Parallel(n_jobs = -1, verbose = 10)(delayed(simulation)(\n",
    "    sample_size = sample_size, num_variables = num_variables, beta_range = beta_range) for _ in range(30))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "d994a4fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x7fd5110a4640>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction = np.concatenate([r[1] for r in result])\n",
    "actual = np.concatenate([r[0] for r in result])\n",
    "RocCurveDisplay.from_predictions(actual, prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "79860606",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7fd550b80760>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = ConfusionMatrixDisplay(confusion_matrix(actual, prediction))\n",
    "ds.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "117a44e1",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}