bootstrap_templates.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "854f842b-b402-4cc5-8559-207c291736a8",
   "metadata": {},
   "source": [
    "# Bootstrap templates"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02722394-c1b0-49b4-b7ee-c143391cac85",
   "metadata": {},
   "source": [
    "## Table of Contents\n",
    "* [Poisson bootstrap for TP, TN, FP, FN template](#1)\n",
    "* [Bootstrap with p value template](#2)\n",
    "* [Poisson Bootstrap with p value template](#3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c17b7d4a-6176-43ef-a082-603986dcbca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import scipy\n",
    "from typing import Callable"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1ac372c-9d64-4e1a-a5fa-c9d7b38031a7",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Poisson bootstrap for TP, TN, FP, FN template <a class=\"anchor\" id=\"1\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "29a08c2b-7b0f-4840-a9ff-5878ef89ce44",
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 100000\n",
    "B = 1000\n",
    "labels = np.random.randint(0, 2, N)  # real labels with two classes 0 and 1\n",
    "predicts = np.random.normal(0.5, 1, N)  # predicted scores from 0 to 1\n",
    "predicts = np.clip(predicts, 0, 1)\n",
    "weights = np.ones(N)  # could be different weights for samples, here all the same"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b692d7c2-b2e8-46c8-9ab4-5152b5b6bcfb",
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = list(zip(labels, predicts, weights))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "1e81b570-9de8-4c1d-9fa6-991db13d244e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def poisson_bootstrap(\n",
    "    data: list[tuple[float, float, float]], \n",
    "    B: int, \n",
    "    threshold: float  # we can pick up most suitable threshold for predictions\n",
    ") -> List[np.array]:\n",
    "    \n",
    "    TP = np.zeros(B)\n",
    "    FP = np.zeros(B)\n",
    "    FN = np.zeros(B)\n",
    "    TN = np.zeros(B)\n",
    "    \n",
    "    for label, predict, weight in data:\n",
    "        predict += np.random.normal(0, 0.05, 1)\n",
    "        predict = np.clip(predict, 0, 1)\n",
    "        predict = int(predict >= threshold)\n",
    "        poisson = np.random.poisson(lam=1, size=B) * weight\n",
    "        \n",
    "        if label == 1 and predict == 1:\n",
    "            TP += poisson\n",
    "        if label == 0 and predict == 0:\n",
    "            TN += poisson\n",
    "        if label == 0 and predict == 1:\n",
    "            FP += poisson\n",
    "        if label == 1 and predict == 0:\n",
    "            FN += poisson\n",
    "            \n",
    "    return TP, FP, FN, TN\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e53ea306-a347-4b50-ad35-4f52d49e11e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([24942699., 25034227., 25123276., 24911285.])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = poisson_bootstrap(data, B, 0.5)\n",
    "np.sum(res, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c5564e45-2738-45fc-aab7-9a97c1ca1ac9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([21094543., 20999662., 28971069., 28943544.])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = poisson_bootstrap(data, B, 0.7)\n",
    "np.sum(res, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "04f6e764-71d1-48d0-91ec-fab8bcc1ff9e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([16919258., 16805860., 33142835., 33127677.])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = poisson_bootstrap(data, B, 0.9)\n",
    "np.sum(res, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de0af190-6c8d-410d-9844-6b9fdc7c2d80",
   "metadata": {},
   "source": [
    "## Bootstrap with p value template <a class=\"anchor\" id=\"2\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "4d46dbeb-617e-4f67-9e89-a745599f93c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_bootstrap(\n",
    "    group_1: np.ndarray,\n",
    "    group_2: np.ndarray,\n",
    "    B: int = 1000, \n",
    "    stat: Callable = np.mean,\n",
    "    bootstrap_conf_level: float = 0.95,\n",
    "    vis: bool = False\n",
    ") -> dict:\n",
    "    \n",
    "    N = max(len(data_column_1), len(data_column_2))\n",
    "    boot_data = []\n",
    "    \n",
    "    for _ in range(B): \n",
    "        samples_1 = np.random.choice(group_1, N)\n",
    "        samples_2 = np.random.choice(group_2, N)\n",
    "        diff = stat(samples_1) - stat(samples_2)\n",
    "        boot_data.append(diff)\n",
    "        \n",
    "    boot_frame = pd.DataFrame(boot_data)\n",
    "    \n",
    "    # p-value calculation\n",
    "    p_1 = scipy.stats.norm.cdf(x = 0, loc = np.mean(boot_data), scale = np.std(boot_data))\n",
    "    p_2 = scipy.stats.norm.cdf(x = 0, loc = -np.mean(boot_data), scale = np.std(boot_data))\n",
    "    p_value = min(p_1, p_2) * 2\n",
    "    \n",
    "    # qiantiles can be used to check if 0 in incide conf interval \n",
    "    left_quant = (1 - bootstrap_conf_level) / 2\n",
    "    right_quant = 1 - ((1 - bootstrap_conf_level) / 2)\n",
    "    quants = boot_frame.quantile([left_quant, right_quant])\n",
    "        \n",
    "    # visualization\n",
    "    if vis:\n",
    "        _, ax = plt.subplots(figsize=(8, 3))\n",
    "        _, _, bars = plt.hist(boot_frame[0], bins = 50)\n",
    "        for bar in bars:\n",
    "            if abs(bar.get_x()) <= quants.iloc[0][0] or abs(bar.get_x()) >= quants.iloc[1][0]:\n",
    "                bar.set_facecolor('red')\n",
    "            else: \n",
    "                bar.set_facecolor('grey')\n",
    "                bar.set_edgecolor('black')\n",
    "\n",
    "        plt.vlines(quants, ymin=0, ymax=50, linestyle='--')\n",
    "        plt.xlabel('boot_data')\n",
    "        plt.ylabel('frequency')\n",
    "        plt.title(\"Histogram of boot_data\")\n",
    "        plt.show()\n",
    "       \n",
    "    return {\"boot_data\": boot_data, \n",
    "            \"quants\": quants, \n",
    "            \"p_value\": p_value}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "13a80923-f8da-4e09-b736-3799312e2693",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAq8AAAE6CAYAAAA4IrvwAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA/O0lEQVR4nO3de1xU1doH8N/mNlxEVEZuKkgJEnhJ9IhoKuYB85ZJvV0QNcuyzJLMPKmVWISKRXTSLMvETseoXtHjOZ0ULLQSNTTUUlDyIGMK4SiK2HCd9f7h6xxHbrNxrvD7fj7zcWbN2nueWbNZPmyeWVsSQggQEREREdkAO0sHQERERERkKCavRERERGQzmLwSERERkc1g8kpERERENoPJKxERERHZDCavRERERGQzmLwSERERkc1g8kpERERENoPJKxERERHZDCavRCRbeno6JEnCwYMHm3x+0qRJ6N27t15b79698eijj8p6ndzcXCQmJuLSpUttC7QD+vzzzxEWFgYXFxdIkoTDhw832a+1z9CUTPG5JiYmQpKkNm27efNmpKWlGS0WIjItJq9EZBZbt27FK6+8Imub3NxcLF++nMmrgc6fP4/p06fj9ttvx44dO7Bv3z4EBwdbOqxGrO1zZfJKZFscLB0AEXUMgwYNsnQIstXV1UGSJDg42MZUefLkSdTV1SE+Ph6jR4+2dDhERCbBM69EZBY3lw1otVokJSWhb9++cHFxQZcuXTBgwAC88847AK79GfjFF18EAAQGBkKSJEiShN27d+u2T0lJQUhICBQKBby8vDBjxgz89ttveq8rhEBycjICAgLg7OyMIUOGIDs7G1FRUYiKitL12717NyRJwt/+9je88MIL6NGjBxQKBX799VecP38ec+fORWhoKDp16gQvLy/cfffd+P777/Ve6/Tp05AkCatXr8aqVavQu3dvuLi4ICoqSpdYvvTSS/Dz84OHhwemTp2K8vJyg8Zv+/btiIyMhKurK9zd3REdHY19+/bpnn/00Udx1113AQAeeughSJKk9/6aU1FRgVmzZqFbt25wc3PD5MmT8Z///KdRv48//hgDBw6Es7MzunXrhqlTp6KgoEB2nK19rob46quvcOedd0KhUCAwMBBvvvlmk/3Wrl2LUaNGwcvLC25ubujfvz9SUlJQV1en6xMVFYWvvvoKJSUlulhuLD9Yvnw5IiIi0K1bN3Tu3Bnh4eHYsGEDhBAGx0tExmUbpxOIyCo1NDSgvr6+Ubsh/7GnpKQgMTERL7/8MkaNGoW6ujoUFhbq/pQ8e/ZsXLx4Ee+++y4yMzPh6+sLAAgNDQUAPP3001i/fj3mzZuHSZMm4fTp03jllVewe/du/PTTT1AqlQCApUuXYsWKFXjyyScRGxuLM2fOYPbs2airq2vyT+qLFy9GZGQk3n//fdjZ2cHLywvnz58HACxbtgw+Pj6oqqrC1q1bERUVhW+++aZRkrh27VoMGDAAa9euxaVLl/DCCy9g8uTJiIiIgKOjIz7++GOUlJRg4cKFmD17NrZv397iWG3evBnTpk1DTEwMPvvsM9TU1CAlJUX3+nfddRdeeeUVDB06FM888wySk5MxZswYdO7cudXP4fHHH0d0dDQ2b96MM2fO4OWXX0ZUVBSOHj2KLl26AABWrFiBJUuW4JFHHsGKFStw4cIFJCYmIjIyEnl5eQgKCjI4ztY+19Z88803mDJlCiIjI5GRkYGGhgakpKTg999/b9T31KlTiIuLQ2BgIJycnHDkyBG88cYbKCwsxMcffwwAeO+99/Dkk0/i1KlT2Lp1a6N9nD59GnPmzIG/vz8AYP/+/Xj22Wdx9uxZvPrqqwbFTERGJoiIZNq4caMA0OItICBAb5uAgAAxc+ZM3eNJkyaJO++8s8XXWb16tQAgiouL9doLCgoEADF37ly99gMHDggAYsmSJUIIIS5evCgUCoV46KGH9Prt27dPABCjR4/WteXk5AgAYtSoUa2+//r6elFXVyfGjh0rpk6dqmsvLi4WAMTAgQNFQ0ODrj0tLU0AEPfee6/efhISEgQAcfny5WZfq6GhQfj5+Yn+/fvr7fPKlSvCy8tLDB8+vNF7+PLLL1t9D9c/wxvjF0KIvXv3CgAiKSlJCCFERUWFcHFxERMmTNDrp1KphEKhEHFxcbLjbO5zNURERITw8/MTGo1G11ZZWSm6desmWvovraGhQdTV1YlPPvlE2Nvbi4sXL+qemzhxYqPjtaV9vPbaa8LT01NotVrZ8RPRrWPZABG12SeffIK8vLxGt+t/vm7J0KFDceTIEcydOxc7d+5EZWWlwa+bk5MDAI1WLxg6dCjuuOMOfPPNNwCunSWrqanBgw8+qNdv2LBhjVZDuO7+++9vsv39999HeHg4nJ2d4eDgAEdHR3zzzTdN/ul8woQJsLP77/R6xx13AAAmTpyo1+96u0qlauadAidOnMC5c+cwffp0vX126tQJ999/P/bv348//vij2e1bM23aNL3Hw4cPR0BAgG6M9+3bB41G02ise/Xqhbvvvls31qaOEwCuXr2KvLw8xMbGwtnZWdfu7u6OyZMnN+qfn5+Pe++9F56enrC3t4ejoyNmzJiBhoYGnDx50qDX/Pbbb/HnP/8ZHh4eun28+uqruHDhgsElH0RkXExeiajN7rjjDgwZMqTRzcPDo9VtFy9ejDfffBP79+/H+PHj4enpibFjxxq0dNOFCxcAQPcn5xv5+fnpnr/+r7e3d6N+TbU1t8/U1FQ8/fTTiIiIwJYtW7B//37k5eXhnnvugUajadS/W7dueo+dnJxabK+urm4ylhvfQ3PvVavVoqKiotntW+Pj49Nk281jaOhYmypO4Fp9rlarbTbmG6lUKowcORJnz57FO++8g++//x55eXlYu3YtADT5ud3sxx9/RExMDADgww8/xN69e5GXl4elS5cavA8iMj7WvBKRRTg4OGDBggVYsGABLl26hF27dmHJkiUYN24czpw5A1dX12a39fT0BACUlpaiZ8+ees+dO3dOV+96vV9T9ZBlZWVNnn1taq3QTz/9FFFRUVi3bp1e+5UrV1p+k0Zw43u92blz52BnZ4euXbu2ef9lZWVNtvXp08eg1795rE0VJwB07doVkiQ1G/ONtm3bhqtXryIzMxMBAQG69ubWvW1KRkYGHB0d8a9//UvvTO+2bdtkx05ExsMzr0RkcV26dMEDDzyAZ555BhcvXsTp06cBAAqFAkDjM1x33303gGtJ5Y3y8vJQUFCAsWPHAgAiIiKgUCjw+eef6/Xbv38/SkpKDI5PkiRdLNcdPXpU71v0ptK3b1/06NEDmzdv1vsi3NWrV7FlyxbdN/vb6u9//7ve49zcXJSUlOi+hBYZGQkXF5dGY/3bb7/h22+/1Y21nDib+1xb4+bmhqFDhyIzM1PvbPWVK1fwz3/+U6/v9V9CbvzchBD48MMPG+1XoVA0Gcv1ZdLs7e11bRqNBn/7299kxU1ExsXklYgsYvLkyVi8eDG2bNmC7777Dn/729+QlpaGgIAA3bfX+/fvDwB45513sG/fPhw8eBBXrlxB37598eSTT+Ldd9/F888/j6ysLKxfvx6TJk1Cr1698PzzzwO49mf6BQsW4IsvvsBTTz2FnTt3YsOGDXjwwQfh6+urV5vZkkmTJiErKwvLli3Dt99+i3Xr1mHcuHEIDAw0zeDcwM7ODikpKTh8+DAmTZqE7du348svv8SYMWNw6dIlrFy58pb2f/DgQcyePRs7d+7ERx99hKlTp6JHjx6YO3cugGu/WLzyyivYvn07ZsyYga+//hqffvopxowZA2dnZyxbtkx2nM19roZ4/fXXUVZWhujoaGzbtg1btmzB2LFj4ebmptcvOjoaTk5OeOSRR/D1119j69atGDduXJOlC/3790d5eTnWrVuHH3/8UVe6MnHiRFRVVSEuLg7Z2dnIyMjAyJEjG/0iQ0RmZuEvjBGRDbr+TfW8vLwmn2/q29s3rzbw1ltvieHDhwulUimcnJyEv7+/ePzxx8Xp06f1tlu8eLHw8/MTdnZ2AoDIyckRQlz75veqVatEcHCwcHR0FEqlUsTHx4szZ87oba/VakVSUpLo2bOncHJyEgMGDBD/+te/xMCBA/W+ad/SN/VramrEwoULRY8ePYSzs7MIDw8X27ZtEzNnztR7n9dXG1i9erXe9s3tu7VxvNG2bdtERESEcHZ2Fm5ubmLs2LFi7969Br1OU66/dlZWlpg+fbro0qWLblWBoqKiRv0/+ugjMWDAAOHk5CQ8PDzElClTxLFjx9oUpxDNf66G2L59uy4Wf39/sXLlSrFs2bJGqw3885//FAMHDhTOzs6iR48e4sUXXxRff/11o9e7ePGieOCBB0SXLl2EJEl6+/n4449F3759hUKhELfddptYsWKF2LBhQ5tXSyCiWycJwZWWiahjKS4uRkhICJYtW4YlS5ZYOhwiIpKBySsRtWtHjhzBZ599huHDh6Nz5844ceIEUlJSUFlZiV9++aXZVQeIiMg6cbUBImrX3NzccPDgQWzYsAGXLl2Ch4cHoqKi8MYbbzBxtSJarRZarbbFPg4O/C+LiHjmlYiIrEBiYiKWL1/eYp/i4uJmLy5BRB0Hk1ciIrK4c+fO4dy5cy32GTBggO7CDkTUcTF5JSIiIiKbwXVeiYiIiMhmtPvqd61Wi3PnzsHd3b3Jyz4SERERkWUJIXDlyhX4+fm1egGZdp+8njt3Dr169bJ0GERERETUijNnzqBnz54t9mn3yau7uzuAa4PRuXNnC0dDRERERDerrKxEr169dHlbS9p98nq9VKBz585MXomIiIismCElnvzCFhERERHZDCavRERERGQzmLwSERERkc1g8kpERERENoPJKxERERHZDCavRERERGQzmLwSERERkc1o9+u8EhF1dCqVCmq12uD+NTU1UCgUBvVVKpXw9/dva2hERLIxeSUiasdUKhVCQkKg0WgM3kaSJAghDOrr4uKCwsJCJrBEZDZMXomI2jG1Wg2NRoPY2FgolcpW+xcVFSEnJ8eg/mq1GpmZmVCr1UxeichsmLwSEXUASqUSfn5+rfa7Xl5gaH8iInNj8kpEZCXk1Kay1pSIOiomr0REVkBubSprTYmoo2LySkRkBeTUprLWlIg6MiavRERWhLWmREQt40UKiIiIiMhmMHklIiIiIpvB5JWIiIiIbAZrXomI6JYUFBQY1I/LexGRMTB5JSKiNqmqqoIkSYiPjzeoP5f3IiJjYPJKRERtUl1dDSEEl/ciIrNi8kpERLeEy3sRkTnxC1tEREREZDOYvBIRERGRzWDySkREREQ2g8krEREREdkMfmGLiMhEVCoV1Gq1QX0NXSuViKijs2jympiYiOXLl+u1eXt7o6ysDAAghMDy5cuxfv16VFRUICIiAmvXrkVYWJglwiUiMphKpUJISAg0Go2lQyEialcsfuY1LCwMu3bt0j22t7fX3U9JSUFqairS09MRHByMpKQkREdH48SJE3B3d7dEuEREBlGr1dBoNAatgQoARUVFyMnJMUNkRES2zeLJq4ODA3x8fBq1CyGQlpaGpUuXIjY2FgCwadMmeHt7Y/PmzZgzZ465QyUiks3QNVANLS8gIuroLJ68FhUVwc/PDwqFAhEREUhOTsZtt92G4uJilJWVISYmRtdXoVBg9OjRyM3NbTZ5rampQU1Nje5xZWWlyd8DEXUchtaxsoaViMg0LJq8RkRE4JNPPkFwcDB+//13JCUlYfjw4Th27Jiu7tXb21tvG29vb5SUlDS7zxUrVjSqoyUiMgbWsRIRWZ5Fk9fx48fr7vfv3x+RkZG4/fbbsWnTJgwbNgwAIEmS3jZCiEZtN1q8eDEWLFige1xZWYlevXoZOXKitqlr0OKzH1UAgEeG+sPRnqvV2RI5daysYSUyHOdGksPiZQM3cnNzQ//+/VFUVIT77rsPAFBWVgZfX19dn/Ly8kZnY2+kUCigUChMHSpRm9Q1aPHqP44BAB4Y3JMTtI0ypI6VNaxEhuPcSHJYVfJaU1ODgoICjBw5EoGBgfDx8UF2djYGDRoEAKitrcWePXuwatUqC0dKRGR5htTVWlvtraHxKJVK+Pv7mzgaIrJFFk1eFy5ciMmTJ8Pf3x/l5eVISkpCZWUlZs6cCUmSkJCQgOTkZAQFBSEoKAjJyclwdXVFXFycJcMmIrKoqqoqSJKE+Ph4S4diMLkxu7i4oLCwkAksETVi0eT1t99+wyOPPAK1Wo3u3btj2LBh2L9/PwICAgAAixYtgkajwdy5c3UXKcjKyuIar0TUoVVXV0MIYVO1t3JiVqvVyMzMhFqtZvJKRI1YNHnNyMho8XlJkpCYmIjExETzBEREZENssfbW0HVviYiaw4poIiIiIrIZTF6JiIiIyGZY1WoDRO2dk70dPn50iO4+ERFxbiR5mLwSmZGDvR3uDml+nWIioo6IcyPJwV9viIiIiMhm8MwrkRnVNWixLf8sAOC+QT14FRkiInBuJHmYvBKZUV2DFi/+71EAwMQBvpygiYjAuZHk4dFBRERERDaDySsRERER2Qwmr0RERERkM5i8EhEREZHNYPJKRERERDaDySsRERER2QwulUVkRk72dlgbF667T0REnBtJHiavRGbkYG+HiQN8LR0GEZFV4dxIcvDXGyIiIiKyGTzzSmRG9Q1a7Dz2OwBgXJg3HPjnMSIizo0kC5NXIjOqbdDimc0/AQCOvzaOEzQRETg3kjw8OoiIiIjIZjB5JSIiIiKbweSViIiIiGwGk1ciIiIishlMXomIiIjIZjB5JSIiIiKbwaWyiMzI0d4Oqx8YoLtPREScG0keJq9EZuRob4f/GdLL0mEQEVkVzo0kB3+9ISIiIiKbwTOvRGZU36DFd0XnAQCjgrrzKjJERODcSPIweSUyo9oGLR5LPwiAl0CUS6VSQa1WG9xfqVTC39/fhBGRqRUUFBjcl5+3bePcSHJYTfK6YsUKLFmyBPPnz0daWhoAQAiB5cuXY/369aioqEBERATWrl2LsLAwywZLRGalUqkQEhICjUZj8DYuLi4oLCxkQmODqqqqIEkS4uPjDd6GnzdRx2EVyWteXh7Wr1+PAQMG6LWnpKQgNTUV6enpCA4ORlJSEqKjo3HixAm4u7tbKFoiMje1Wg2NRoPY2FgolUqD+mdmZkKtVjOZsUHV1dUQQvDzJqImWTx5raqqwrRp0/Dhhx8iKSlJ1y6EQFpaGpYuXYrY2FgAwKZNm+Dt7Y3Nmzdjzpw5Te6vpqYGNTU1useVlZWmfQNEZDZKpRJ+fn5G36+hJQly/oxNt07u523o58MSAyLbZvHk9ZlnnsHEiRPx5z//WS95LS4uRllZGWJiYnRtCoUCo0ePRm5ubrPJ64oVK7B8+XKTx01E7UNbShLIusgtM2CJAZFts2jympGRgZ9++gl5eXmNnisrKwMAeHt767V7e3ujpKSk2X0uXrwYCxYs0D2urKxEr15cO46ImianJKGoqAg5OTlmiowMJafMgCUGRLbPYsnrmTNnMH/+fGRlZcHZ2bnZfpIk6T0WQjRqu5FCoYBCoTBanETUMRjyJ2o5qx2Q+ZmqrISIrIvFktdDhw6hvLwcgwcP1rU1NDTgu+++w5o1a3DixAkA187A+vr66vqUl5c3OhtLZCsc7e3w2pQw3X0iIuLcSPJYLHkdO3Ysfv75Z722WbNmISQkBH/5y19w2223wcfHB9nZ2Rg0aBAAoLa2Fnv27MGqVassETLRLXO0t8OMyN6WDoOIyKpwbiQ5LJa8uru7o1+/fnptbm5u8PT01LUnJCQgOTkZQUFBCAoKQnJyMlxdXREXF2eJkImIiIjIwiy+2kBLFi1aBI1Gg7lz5+ouUpCVlcU1XslmNWgFfiy+CAAYGtgN9nbN128TEXUUnBtJDqtKXnfv3q33WJIkJCYmIjEx0SLxEBlbTX0DHvlwP4Brl0B0dbKqH0GiDoNrwloXzo0kB48OIiLqMLgmLJHtY/JKREQdBteEJbJ9spPX4uJiBAYGmiIWIiIis+CasES2S/Zian369MGYMWPw6aeforq62hQxERERERE1SXbyeuTIEQwaNAgvvPACfHx8MGfOHPz444+miI2IiIiISI/s5LVfv35ITU3F2bNnsXHjRpSVleGuu+5CWFgYUlNTcf78eVPESUREREQkP3m9zsHBAVOnTsUXX3yBVatW4dSpU1i4cCF69uyJGTNmoLS01JhxErULDnZ2WDw+BIvHh8DBjpdAJCICODeSPG1ebeDgwYP4+OOPkZGRATc3NyxcuBCPP/44zp07h1dffRVTpkxhOQHRTZwc7DBn9O2WDoOIZOCasKbHuZHkkJ28pqamYuPGjThx4gQmTJiATz75BBMmTIDd//+mFBgYiA8++AAhISFGD5aIiMhcuCYskXWSnbyuW7cOjz32GGbNmgUfH58m+/j7+2PDhg23HBxRe9OgFfjl7GUAQL8eHrwEIpEV45qw5sO5keSQnbwWFRW12sfJyQkzZ85sU0BE7VlNfQOmrN0LgJdAJLIVXBPW9Dg3khyyq6I3btyIL7/8slH7l19+iU2bNhklKCIiIiKipshOXleuXNnkn0+8vLyQnJxslKCIiIiIiJoiO3ktKSlp8vKwAQEBUKlURgmKiIiIiKgpspNXLy8vHD16tFH7kSNH4OnpaZSgiIiIiIiaIjt5ffjhh/Hcc88hJycHDQ0NaGhowLfffov58+fj4YcfNkWMREREREQA2rDaQFJSEkpKSjB27Fg4OFzbXKvVYsaMGax5JSIiIiKTkp28Ojk54fPPP8frr7+OI0eOwMXFBf3790dAQIAp4iNqVxzs7DB/bJDuPhERcW4kedq8kFpwcDCCg4ONGQtRu+fkYIfno/lzQ0R0I86NJIfs5LWhoQHp6en45ptvUF5eDq1Wq/f8t99+a7TgiIhuhSHXpDf0uvVERGQdZCev8+fPR3p6OiZOnIh+/fpBkngJNyJDabUCv56vAgD06d4JdrwEoknIvSY9EVkW50aSQ3bympGRgS+++AITJkwwRTxE7Vp1fQNi3v4OAC+BaEpyrklfVFSEnJwcM0VGRE3h3EhytOkLW3369DFFLERERmXINenVarWZoiEiImOQ/ZW+F154Ae+88w6EEKaIh4iIiIioWbLPvP7www/IycnB119/jbCwMDg6Ouo9n5mZabTgiIiIiIhuJDt57dKlC6ZOnWqKWIiIiIiIWiQ7ed24caMp4iAiIiIialWbvs5XX1+P3bt349SpU4iLi4O7uzvOnTuHzp07o1OnTsaOkYhsiEqlMvhLUEqlEv7+/iaOiIiI2hPZyWtJSQnuueceqFQq1NTUIDo6Gu7u7khJSUF1dTXef/99g/e1bt06rFu3DqdPnwYAhIWF4dVXX8X48eMBAEIILF++HOvXr0dFRQUiIiKwdu1ahIWFyQ2byCo42NnhyVG36e63NyqVCiEhIdBoNAb1d3FxQWFhIRNYog6uvc+NZFxtukjBkCFDcOTIEXh6eurap06ditmzZ8vaV8+ePbFy5Urd0lubNm3ClClTkJ+fj7CwMKSkpCA1NRXp6ekIDg5GUlISoqOjceLECbi7u8sNncjinBzssGTCHZYOw2TUajU0Go1B66uq1WpkZmZCrVYzeSXq4Nr73EjG1abVBvbu3QsnJye99oCAAJw9e1bWviZPnqz3+I033sC6deuwf/9+hIaGIi0tDUuXLkVsbCyAa8mtt7c3Nm/ejDlz5sgNnYjMxJD1VYmIiNpC9rl5rVaLhoaGRu2//fbbLZ0NbWhoQEZGBq5evYrIyEgUFxejrKwMMTExuj4KhQKjR49Gbm5us/upqalBZWWl3o3IWmi1Amcu/oEzF/+AVsu1komIAM6NJI/s5DU6OhppaWm6x5IkoaqqCsuWLWvTJWN//vlndOrUCQqFAk899RS2bt2K0NBQlJWVAQC8vb31+nt7e+uea8qKFSvg4eGhu/Xq1Ut2TESmUl3fgJEpORiZkoPq+sa/BBIRdUScG0kO2cnr22+/jT179iA0NBTV1dWIi4tD7969cfbsWaxatUp2AH379sXhw4exf/9+PP3005g5cyaOHz+ue16SJL3+QohGbTdavHgxLl++rLudOXNGdkxEREREZJ1k17z6+fnh8OHD+Oyzz/DTTz9Bq9Xi8ccfx7Rp0+Di4iI7ACcnJ90XtoYMGYK8vDy88847+Mtf/gIAKCsrg6+vr65/eXl5o7OxN1IoFFAoFLLjICIiIiLr16Z1Xl1cXPDYY4/hscceM3Y8EEKgpqYGgYGB8PHxQXZ2NgYNGgQAqK2txZ49e9p0hpeIiIiIbJ/s5PWTTz5p8fkZM2YYvK8lS5Zg/Pjx6NWrF65cuYKMjAzs3r0bO3bsgCRJSEhIQHJyMoKCghAUFITk5GS4uroiLi5ObthERERE1A60aZ3XG9XV1eGPP/6Ak5MTXF1dZSWvv//+O6ZPn47S0lJ4eHhgwIAB2LFjB6KjowEAixYtgkajwdy5c3UXKcjKyuIar0REREQdlOzktaKiolFbUVERnn76abz44ouy9rVhw4YWn5ckCYmJiUhMTJS1XyIiIiJqn9pU83qzoKAgrFy5EvHx8SgsLDTGLonaJXs7CdOHBejuW4pKpYJarTa4v1Kp5FWwiIxMzs9he/8ZtJa5kWyDUZJXALC3t8e5c+eMtTuidknhYI/X7+tn0RhUKhVCQkKg0WgM3sbFxQWFhYXt+j9PInOS+3PY3n8GrWFuJNshO3ndvn273mMhBEpLS7FmzRqMGDHCaIERkWmo1WpoNBrExsZCqVQa1D8zMxNqtbrd/sdJZG5yfg75M0ikT3byet999+k9liQJ3bt3x91334233nrLWHERtUtCCFy8WgsA6Obm1OIFN0xNqVTCz8/PYq9PRPw5vM6a5kayfrKTV61Wa4o4iDoETV0DBiftAgAcf20cXJ2MVrlDRGSzODeSHLIvD0tEREREZCmyf7VZsGCBwX1TU1Pl7p6IiIiIqFmyk9f8/Hz89NNPqK+vR9++fQEAJ0+ehL29PcLDw3X9WK9CREQdTUFBgVH7EVFjspPXyZMnw93dHZs2bULXrl0BXLtwwaxZszBy5Ei88MILRg+SiIjImlVVVUGSJMTHx1s6FKJ2T3by+tZbbyErK0uXuAJA165dkZSUhJiYGCavRETU4VRXV0MIYfASdEVFRcjJyTFDZETtj+zktbKyEr///jvCwsL02svLy3HlyhWjBUZERGRrDF36Ss4V7ohIn+zkderUqZg1axbeeustDBs2DACwf/9+vPjii4iNjTV6gETtib2dhPvDe+ruExER50aSR3by+v7772PhwoWIj49HXV3dtZ04OODxxx/H6tWrjR4gUXuicLDHWw8OtHQYRERWhXMjySE7eXV1dcV7772H1atX49SpUxBCoE+fPnBzczNFfEREREREOm2+SEFpaSlKS0sRHBwMNzc3CCGMGRdRuySEwB+19fijtp4/M0RE/49zI8khO3m9cOECxo4di+DgYEyYMAGlpaUAgNmzZ3OlAaJWaOoaEPrqToS+uhOaugZLh0NEZBU4N5IcspPX559/Ho6OjlCpVHB1ddW1P/TQQ9ixY4dRgyMiIiIiupHsmtesrCzs3LkTPXv21GsPCgpCSUmJ0QIjIiIiIrqZ7DOvV69e1Tvjep1arYZCoTBKUERERERETZGdvI4aNQqffPKJ7rEkSdBqtVi9ejXGjBlj1OCIiIiIiG4ku2xg9erViIqKwsGDB1FbW4tFixbh2LFjuHjxIvbu3WuKGImIiIiIALThzGtoaCiOHj2KoUOHIjo6GlevXkVsbCzy8/Nx++23myJGIiIiIiIAMs+81tXVISYmBh988AGWL19uqpiI2i07ScKE/j66+0RExLmR5JGVvDo6OuKXX36BxAOLqE2cHe3x3rTBlg6DiGxQQUGBQf2USiX8/f1NHI1xcW4kOWTXvM6YMQMbNmzAypUrTREPERER3aCqqgqSJCE+Pt6g/i4uLigsLLS5BJbIULKT19raWnz00UfIzs7GkCFD4Obmpvd8amqq0YIjIiLq6KqrqyGEQGxsLJRKZYt91Wo1MjMzoVarmbxSu2VQ8nr06FH069cPdnZ2+OWXXxAeHg4AOHnypF4/lhMQteyP2nqEvroTAHD8tXFwdZL9+yMRdVBKpRJ+fn6WDsMkODeSHAYdHYMGDUJpaSm8vLxQUlKCvLw8eHp6mjo2IiIiagND62MB26yRpY7NoOS1S5cuKC4uhpeXF06fPg2tVmuUF1+xYgUyMzNRWFgIFxcXDB8+HKtWrULfvn11fYQQWL58OdavX4+KigpERERg7dq1CAsLM0oMRERE7YXc+liANbJkewxKXu+//36MHj0avr6+kCQJQ4YMgb29fZN9//Of/xj84nv27MEzzzyDP/3pT6ivr8fSpUsRExOD48eP62ppU1JSkJqaivT0dAQHByMpKQnR0dE4ceIE3N3dDX4tIiKi9k5OfSzAGlmyTQYlr+vXr0dsbCx+/fVXPPfcc3jiiSeMkjju2LFD7/HGjRvh5eWFQ4cOYdSoURBCIC0tDUuXLkVsbCwAYNOmTfD29sbmzZsxZ86cW46BiIiovWnP9bFEBldE33PPPQCAQ4cOYf78+SY563n58mUAQLdu3QAAxcXFKCsrQ0xMjK6PQqHA6NGjkZub22TyWlNTg5qaGt3jyspKo8dJRMZjSG2enPo9IiJq32R/nW/jxo2miANCCCxYsAB33XUX+vXrBwAoKysDAHh7e+v19fb2RklJSZP7WbFiBa/+RWQD2lKbR0REZDVrUcybNw9Hjx7FDz/80Oi5m5fgEkI0uyzX4sWLsWDBAt3jyspK9OrVy7jBErWRnSRhTN/uuvsdmZzavKKiIuTk5JgpMiIyN86NJIdVJK/PPvsstm/fju+++w49e/bUtfv4XLvOcVlZGXx9fXXt5eXljc7GXqdQKKBQKEwbMFEbOTvaY+OsoZYOw6oYUpunVqvNFA0RWQLnRpLDzpIvLoTAvHnzkJmZiW+//RaBgYF6zwcGBsLHxwfZ2dm6ttraWuzZswfDhw83d7hEREREZGEWPfP6zDPPYPPmzfjHP/4Bd3d3XY2rh4cHXFxcIEkSEhISkJycjKCgIAQFBSE5ORmurq6Ii4uzZOhEREREZAEWTV7XrVsHAIiKitJr37hxIx599FEAwKJFi6DRaDB37lzdRQqysrK4xivZpD9q6zH49V0AgEOv/JmXQCQiAudGkseiR4cQotU+kiQhMTERiYmJpg+IyAw0dQ2WDoGIyOpwbiRDWbTmlYiIiIhIDiavRERERGQzmLwSERERkc1g8kpERERENoPJKxERERHZDK5FQWRGdpKEiMBuuvtERMS5keRh8kpkRs6O9vh8TqSlwyAisiqcG0kOlg0QERERkc1g8kpERERENoPJK5EZ/VFbj/DXsxH+ejb+qK23dDhERFaBcyPJwZpXIjO7eLXW0iEQEVkdzo1kKJ55JSIiIiKbwTOvREREHVxBQYFB/ZRKJfz9/U0cDVHLmLwSERF1UFVVVZAkCfHx8Qb1d3FxQWFhIRNYsigmr0RERB1UdXU1hBCIjY2FUqlssa9arUZmZibUajWTV7IoJq9EREQdnFKphJ+fn6XDIDIIk1ciM7KTJAzo6aG73xqVSgW1Wm3QvlmLRkTmYIr6WLlzI3VsTF6JzMjZ0R7b591lUF+VSoWQkBBoNBqD+rMWjYhMyZT1sXLmRiImr0RWSq1WQ6PRsBaNiKwC62PJWjB5JbJyrEUjImvCOYksjckrkRmdPHUa0/5eCAD46z3doXBovrbL0LoyczEkHmuLmYhsg6a2AX9O3QMA2LVgNFyc7C0cEVkzJq9EZqJSqXDnoEHwmvspAGDEXSMg6mosHFXr5Na5ERHJJSBw9pJGd5+oJUxeicxErVajuvq/X76aNesxOKCh2f5FRUXIyckxR2gtklPnZi0xExFR+8XklchCfH194Shpm33e0CWyzMWQOjdri5mIiNofO0sHQERERERkKCavRERERGQzmLwSERERkc1gzSuROQnAraESjg6O4AUQiYiukSAhyKuT7j5RS5i8EpmRqK/BXZW7ucA3EdENXJzskb1gtKXDIBth0bKB7777DpMnT4afnx8kScK2bdv0nhdCIDExEX5+fnBxcUFUVBSOHTtmmWCJiIiIyOIsmrxevXoVAwcOxJo1a5p8PiUlBampqVizZg3y8vLg4+OD6OhoXLlyxcyREhEREZE1sGjZwPjx4zF+/PgmnxNCIC0tDUuXLkVsbCwAYNOmTfD29sbmzZsxZ86cJrerqalBTc1/r1pUWVlp/MCJ2khyUOCHzlFwrHbEZEUBHFpY55WIqKPQ1Dbg3jU/AAC2z7uLl4elFlntagPFxcUoKytDTEyMrk2hUGD06NHIzc1tdrsVK1bAw8NDd+vVq5c5wiUyjARcte+MS8KFF0AkIvp/AgJF5VUoKq/i5WGpVVabvJaVlQEAvL299dq9vb11zzVl8eLFuHz5su525swZk8ZJREREROZj9asNSJL+khlCiEZtN1IoFFAoFKYOi4iIiIgswGqTVx8fHwDXzsD6+vrq2svLyxudjSUiIiLrU1BQYFA/N49uJo6E2hOrTV4DAwPh4+OD7OxsDBo0CABQW1uLPXv2YNWqVRaOjoiIiJpTVVUFSZIQHx9vUH8X9y7wmvupiaOi9sKiyWtVVRV+/fVX3ePi4mIcPnwY3bp1g7+/PxISEpCcnIygoCAEBQUhOTkZrq6uiIuLs2DURERE1JLq6moIIRAbGwulUtliX7Vaja3//MpMkVF7YNHk9eDBgxgzZozu8YIFCwAAM2fORHp6OhYtWgSNRoO5c+eioqICERERyMrKgru7u6VCJro1AnBu+AMODva8ACIRtXtKpdKwKwoKoLurPZycnHh5WGqVRZPXqKgoCNH8khiSJCExMRGJiYnmC4oIgEqlglqtNqivUqmEv7+/QX1FfQ1GV+4y2eVhDakvM7QGjYjIXER9DT6Y5IXw8HBLh0I2wGprXoksRaVSISQkBBqNxqD+Li4uKCwsNDiBNQW59WVERES2iskr0U3UajU0Go3BtVqZmZlQq9UWTV7l1JcVFRUhJyfHTJEREREZF5NXomYYXKslg+TghH3uI+FU7YTxikI4SMa9kowhMRtaDkFEZC6SgxNezFbDLfcHfDEnEs6OvDwsNY/JK5ERGFxrKkmodOgKCEBAAngZRCIiQJJwqqIOqLgMbQvfhSECmLwS3RK5taaSI6/+RkREdCuYvBLdArm1prt/yDVTZERERO0Tk1ciI2CtKRERkXnYWToAIiIiIiJD8cwrEREREQBIMq7uxS+WWQyTVyIzc9TWwM6Of/QgIrpRZ4UdHByYllDreJQQmZGoq8Hdl3ea7PKwRES2SNTVIH2KNy8PSwZh8kpEREQWZ8h62dcplUqLXtWQLIvJKxEREVmM3PWyAZmXd2FtarvD5JXIjCQHJ/zYaTgUNQpEO500+uVhiYhsTXV1NWDviLCn18LR0RGDq/bDHtpm+6vVaiAz04wRkrVh8kpkTpKECkcloOXlYYmIdCQJVa6+AAAfXz84Ss0nr0RMXslmqVQqWQv/s0aKiIjI9jF5JZukUqkQEhICjUZj8DYuLi4oLCxkAktE1JHIWbuVbAKTV7JJarUaGo0GsbGxUCqVBvXPzMyEWq1m8kpERGTDmLySTVMqlVwzlYiIqANh8kodiiHrCMpZa5CIiIjMi8krdQhtWUfQVOxFPSTWYBER6eHcSIZi8kodQnV1NYQQBtXIFhUVIScnxyRxiLoa/PnSv1nqQER0A86NJAeTV+pQDKmRlbP8FhEREZmXnaUDICIiIiIyFJNXInOyd8ShThHIrumDesHaLiIiALY5N0qS6W7UIpYNEJmRZGcHtaM3Lw9LRHQDzo0kB8+8EhEREZHN4JlXahOVSmXwF5uUSiWvakVERERGYRPJ63vvvYfVq1ejtLQUYWFhSEtLw8iRIy0dVvPk1KsI2/vTiEqlQkhICDQajUH9XVxcUFhYyASWiIjIEKbMI9pBjmL1yevnn3+OhIQEvPfeexgxYgQ++OADjB8/HsePH2cyZCFqtRoajcagNVPVajUyMzOhVqv5eREREdEts/rkNTU1FY8//jhmz54NAEhLS8POnTuxbt06rFixwsLRdWyGrJlKREREZExWnbzW1tbi0KFDeOmll/TaY2JikJub2+Q2NTU1qKmp0T2+fPkyAKCystJ0gd4KK4mrrKwMZWVlBvU9ceIEAKC0tBS1tbUt9r1w4QIA4NChQ6iqqmp133Z2dtBqtUaNAQDOnz9vcH9T9hVCQFvzBwCg5PcSOKDBqmO2ljgYc8eKgzF3rDjkzo0XLlyAdfzPaSVMmUeYMUe5nqcJQ0oVhBU7e/asACD27t2r1/7GG2+I4ODgJrdZtmyZwLU1NnjjjTfeeOONN954s6HbmTNnWs0PrfrM63XSTcXFQohGbdctXrwYCxYs0D3WarW4ePEiPD09ceXKFfTq1QtnzpxB586dTRozXfstiuNtPhxv8+J4mxfH27w43ubF8b6W2125csWgckSrTl6VSiXs7e0b/Tm7vLwc3t7eTW6jUCigUCj02rp06QLgv0lw586dO+zBYQkcb/PieJsXx9u8ON7mxfE2r44+3h4eHgb1s+qLFDg5OWHw4MHIzs7Wa8/Ozsbw4cMtFBURERERWYpVn3kFgAULFmD69OkYMmQIIiMjsX79eqhUKjz11FOWDo2IiIiIzMzqk9eHHnoIFy5cwGuvvYbS0lL069cP//73vxEQECB7XwqFAsuWLWtUVkCmwfE2L463eXG8zYvjbV4cb/PieMsjCWGll08gIiIiIrqJVde8EhERERHdiMkrEREREdkMJq9EREREZDOYvBIRERGRzWhXyWtFRQWmT58ODw8PeHh4YPr06bh06VKL22RmZmLcuHFQKpWQJAmHDx9u1CcqKgqSJOndHn74YdO8CRtiqvGuqanBs88+C6VSCTc3N9x777347bffTPMmbEhbxlsIgcTERPj5+cHFxQVRUVE4duyYXh8e39e89957CAwMhLOzMwYPHozvv/++xf579uzB4MGD4ezsjNtuuw3vv/9+oz5btmxBaGgoFAoFQkNDsXXrVlOFb3OMPd7p6emNjmNJklBdXW3Kt2Ez5Ix3aWkp4uLi0LdvX9jZ2SEhIaHJfjy+m2fs8ebxra9dJa9xcXE4fPgwduzYgR07duDw4cOYPn16i9tcvXoVI0aMwMqVK1vs98QTT6C0tFR3++CDD4wZuk0y1XgnJCRg69atyMjIwA8//ICqqipMmjQJDQ0Nxn4LNqUt452SkoLU1FSsWbMGeXl58PHxQXR0NK5cuaLXr6Mf359//jkSEhKwdOlS5OfnY+TIkRg/fjxUKlWT/YuLizFhwgSMHDkS+fn5WLJkCZ577jls2bJF12ffvn146KGHMH36dBw5cgTTp0/Hgw8+iAMHDpjrbVktU4w3cO3qRDcex6WlpXB2djbHW7Jqcse7pqYG3bt3x9KlSzFw4MAm+/D4bp4pxhvg8a1HtBPHjx8XAMT+/ft1bfv27RMARGFhYavbFxcXCwAiPz+/0XOjR48W8+fPN2K0ts9U433p0iXh6OgoMjIydG1nz54VdnZ2YseOHUaL39a0Zby1Wq3w8fERK1eu1LVVV1cLDw8P8f777+vaeHwLMXToUPHUU0/ptYWEhIiXXnqpyf6LFi0SISEhem1z5swRw4YN0z1+8MEHxT333KPXZ9y4ceLhhx82UtS2yxTjvXHjRuHh4WH0WNsDueN9o+bmBx7fzTPFePP41tduzrzu27cPHh4eiIiI0LUNGzYMHh4eyM3NveX9//3vf4dSqURYWBgWLlzY6MxVR2Oq8T506BDq6uoQExOja/Pz80O/fv2M8jnaqraMd3FxMcrKyvTGUqFQYPTo0Y226cjHd21tLQ4dOqQ3TgAQExPT7Nju27evUf9x48bh4MGDqKura7FPRz6OAdONNwBUVVUhICAAPXv2xKRJk5Cfn2/8N2Bj2jLehuDx3TRTjTfA4/tGVn+FLUOVlZXBy8urUbuXlxfKyspuad/Tpk1DYGAgfHx88Msvv2Dx4sU4cuQIsrOzb2m/tsxU411WVgYnJyd07dpVr93b2/uWP0db1pbxvt7u7e2t1+7t7Y2SkhLd445+fKvVajQ0NDQ5Ti2NbVP96+vroVar4evr22yfjnwcA6Yb75CQEKSnp6N///6orKzEO++8gxEjRuDIkSMICgoy2fuxdm0Zb0Pw+G6aqcabx7c+q09eExMTsXz58hb75OXlAQAkSWr0nBCiyXY5nnjiCd39fv36ISgoCEOGDMFPP/2E8PDwW9q3tbGG8W6KqfZraeYY75ufv3mbjnR8t6S1cTKk/83tcvfZkRh7vIcNG4Zhw4bpnh8xYgTCw8Px7rvv4q9//auxwrZZpjgWeXw3z9hjw+Nbn9Unr/PmzWv1m8+9e/fG0aNH8fvvvzd67vz5841+A7pV4eHhcHR0RFFRUbv7z93S4+3j44Pa2lpUVFTonX0tLy/H8OHD27xfa2XK8fbx8QFw7QyJr6+vrr28vLzFz6g9H99NUSqVsLe3b3RWpKVx8vHxabK/g4MDPD09W+xj7PnI1phqvG9mZ2eHP/3pTygqKjJO4DaqLeNtCB7fTTPVeN+sox/fVl/zqlQqERIS0uLN2dkZkZGRuHz5Mn788UfdtgcOHMDly5eNnvQcO3YMdXV1eglBe2Hp8R48eDAcHR31/mRdWlqKX375pV0mr6Yc7+ulADeOZW1tLfbs2dPiWLbn47spTk5OGDx4cKMyiezs7GbHKTIyslH/rKwsDBkyBI6Oji32aY/HsRymGu+bCSFw+PDhDnMcN6ct420IHt9NM9V436zDH98W+JKYydxzzz1iwIABYt++fWLfvn2if//+YtKkSXp9+vbtKzIzM3WPL1y4IPLz88VXX30lAIiMjAyRn58vSktLhRBC/Prrr2L58uUiLy9PFBcXi6+++kqEhISIQYMGifr6erO+P2tjivEWQoinnnpK9OzZU+zatUv89NNP4u677xYDBw7keLdhvFeuXCk8PDxEZmam+Pnnn8UjjzwifH19RWVlpRCCx/d1GRkZwtHRUWzYsEEcP35cJCQkCDc3N3H69GkhhBAvvfSSmD59uq7/f/7zH+Hq6iqef/55cfz4cbFhwwbh6Ogo/vd//1fXZ+/evcLe3l6sXLlSFBQUiJUrVwoHBwe9FSM6KlOMd2JiotixY4c4deqUyM/PF7NmzRIODg7iwIEDZn9/1kbueAshRH5+vsjPzxeDBw8WcXFxIj8/Xxw7dkz3PI/v5plivHl862tXyeuFCxfEtGnThLu7u3B3dxfTpk0TFRUVen0AiI0bN+oeb9y4UQBodFu2bJkQQgiVSiVGjRolunXrJpycnMTtt98unnvuOXHhwgXzvTErZYrxFkIIjUYj5s2bJ7p16yZcXFzEpEmThEqlMs+bsmJtGW+tViuWLVsmfHx8hEKhEKNGjRI///yz7nke3/+1du1aERAQIJycnER4eLjYs2eP7rmZM2eK0aNH6/XfvXu3GDRokHBychK9e/cW69ata7TPL7/8UvTt21c4OjqKkJAQsWXLFlO/DZth7PFOSEgQ/v7+wsnJSXTv3l3ExMSI3Nxcc7wVmyB3vJuapwMCAvT68PhunrHHm8e3PkmI/696JyIiIiKyclZf80pEREREdB2TVyIiIiKyGUxeiYiIiMhmMHklIiIiIpvB5JWIiIiIbAaTVyIiIiKyGUxeiYiIiMhmMHklIiIiIpvB5JWISIaoqCgkJCRYOoxWpaeno0uXLpYOg4jI6Ji8EhFZmd69eyMtLa3DvC4RkRxMXomIiIjIZjB5JSKSqb6+HvPmzUOXLl3g6emJl19+GUIIAEBFRQVmzJiBrl27wtXVFePHj0dRUZHe9lu2bEFYWBgUCgV69+6Nt956S/dcVFQUSkpK8Pzzz0OSJEiSZFBM6enp8Pf3h6urK6ZOnYoLFy7oPX/q1ClMmTIF3t7e6NSpE/70pz9h165drb7uhQsX8Mgjj6Bnz55wdXVF//798dlnn7Vp3IiIjIHJKxGRTJs2bYKDgwMOHDiAv/71r3j77bfx0UcfAQAeffRRHDx4ENu3b8e+ffsghMCECRNQV1cHADh06BAefPBBPPzww/j555+RmJiIV155Benp6QCAzMxM9OzZE6+99hpKS0tRWlraajwHDhzAY489hrlz5+Lw4cMYM2YMkpKS9PpUVVVhwoQJ2LVrF/Lz8zFu3DhMnjwZKpWqxdetrq7G4MGD8a9//Qu//PILnnzySUyfPh0HDhww1nASEckiieunC4iIqFVRUVEoLy/HsWPHdGcnX3rpJWzfvh3/+Mc/EBwcjL1792L48OEArp257NWrFzZt2oT/+Z//wbRp03D+/HlkZWXp9rlo0SJ89dVXOHbsGIBrtacJCQkGfzEsLi4OFRUV+Prrr3VtDz/8MHbs2IFLly41u11YWBiefvppzJs3T9brTpw4EXfccQfefPNNg+IjIjImnnklIpJp2LBhen/Oj4yMRFFREY4fPw4HBwdERETonvP09ETfvn1RUFAAACgoKMCIESP09jdixAgUFRWhoaGhTfEUFBQgMjJSr+3mx1evXsWiRYsQGhqKLl26oFOnTigsLNSdeW1OQ0MD3njjDQwYMACenp7o1KkTsrKyWt2OiMhUHCwdABFReyeE0CW7N96/8flb3X9rXnzxRezcuRNvvvkm+vTpAxcXFzzwwAOora1tcbu33noLb7/9NtLS0tC/f3+4ubkhISGh1e2IiEyFySsRkUz79+9v9DgoKAihoaGor6/HgQMH9MoGTp48iTvuuAMAEBoaih9++EFv+9zcXAQHB8Pe3h4A4OTkJOssbGhoaJMx3ej777/Ho48+iqlTpwK4VgN7+vRpvT5Nve7333+PKVOmID4+HgCg1WpRVFSkez9ERObGsgEiIpnOnDmDBQsW4MSJE/jss8/w7rvvYv78+QgKCsKUKVPwxBNP4IcffsCRI0cQHx+PHj16YMqUKQCAF154Ad988w1ef/11nDx5Eps2bcKaNWuwcOFC3f579+6N7777DmfPnoVarW41nueeew47duxASkoKTp48iTVr1mDHjh16ffr06YPMzEwcPnwYR44cQVxcHLRarV6fpl63T58+yM7ORm5uLgoKCjBnzhyUlZXd6hASEbUZk1ciIplmzJgBjUaDoUOH4plnnsGzzz6LJ598EgCwceNGDB48GJMmTUJkZCSEEPj3v/8NR0dHAEB4eDi++OILZGRkoF+/fnj11Vfx2muv4dFHH9Xt/7XXXsPp06dx++23o3v37q3GM2zYMHz00Ud49913ceeddyIrKwsvv/yyXp+3334bXbt2xfDhwzF58mSMGzcO4eHhen2aet1XXnkF4eHhGDduHKKiouDj44P77rvvFkaPiOjWcLUBIiIiIrIZPPNKRERERDaDySsRkZUbP348OnXq1OQtOTnZ0uEREZkVywaIiKzc2bNnodFomnyuW7du6Natm5kjIiKyHCavRERERGQzWDZARERERDaDySsRERER2Qwmr0RERERkM5i8EhEREZHNYPJKRERERDaDySsRERER2Qwmr0RERERkM/4PRB+qEUS1UrQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.5623642611094166"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "N = 1000\n",
    "sample_1 = np.random.normal(0.5, 1, N)\n",
    "sample_2 = np.random.normal(0.5, 1, N)\n",
    "\n",
    "res = get_bootstrap(sample_1, sample_2, B = 1000, vis=True)\n",
    "res[\"p_value\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "b63c594c-1075-4a9c-9bd8-06eac199aaf0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.04276808180104267"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "N = 1000\n",
    "sample_1 = np.random.normal(0.5, 1, N)\n",
    "sample_2 = np.random.normal(0.5, 1, N) + 0.1\n",
    "\n",
    "res = get_bootstrap(sample_1, sample_2, B = 1000, vis=True)\n",
    "res[\"p_value\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96b98a7a-d1d1-436c-8aeb-48856cac796a",
   "metadata": {},
   "source": [
    "## Poisson Bootstrap with p value template <a class=\"anchor\" id=\"3\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "368672f4-1e78-4ffa-bac5-49e43e50bbd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_poisson_bootstrap(\n",
    "    group_1: np.ndarray,\n",
    "    group_2: np.ndarray,\n",
    "    B: int = 10000,\n",
    "    bootstrap_conf_level: float = 0.95,\n",
    "    vis: bool = False\n",
    ") -> dict:\n",
    "    \n",
    "    A_group = np.zeros(B)\n",
    "    B_group = np.zeros(B)\n",
    "    # separate loop for case where lengths are different\n",
    "    for i in range(len(data_column_1)):\n",
    "        poisson_weights = np.random.poisson(lam=1, size=B)\n",
    "        A_group += poisson_weights * group_1[i]\n",
    "        \n",
    "    for i in range(len(data_column_2)):\n",
    "        poisson_weights = np.random.poisson(lam=1, size=B)\n",
    "        B_group += poisson_weights * group_2[i]\n",
    "\n",
    "    A_group /= N\n",
    "    B_group /= N\n",
    "\n",
    "    boot_data = A_group - B_group\n",
    "        \n",
    "    boot_frame = pd.DataFrame(boot_data)\n",
    "    \n",
    "    # p-value calculation\n",
    "    p_1 = scipy.stats.norm.cdf(x = 0, loc = np.mean(boot_data), scale = np.std(boot_data))\n",
    "    p_2 = scipy.stats.norm.cdf(x = 0, loc = -np.mean(boot_data), scale = np.std(boot_data))\n",
    "    p_value = min(p_1, p_2) * 2\n",
    "    \n",
    "    # qiantiles can be used to check if 0 in incide conf interval \n",
    "    left_quant = (1 - bootstrap_conf_level) / 2\n",
    "    right_quant = 1 - ((1 - bootstrap_conf_level) / 2)\n",
    "    quants = boot_frame.quantile([left_quant, right_quant])\n",
    "        \n",
    "    # visualization\n",
    "    if vis:\n",
    "        _, ax = plt.subplots(figsize=(8, 3))\n",
    "        _, _, bars = plt.hist(boot_frame[0], bins = 50)\n",
    "        for bar in bars:\n",
    "            if abs(bar.get_x()) <= quants.iloc[0][0] or abs(bar.get_x()) >= quants.iloc[1][0]:\n",
    "                bar.set_facecolor('red')\n",
    "            else: \n",
    "                bar.set_facecolor('grey')\n",
    "                bar.set_edgecolor('black')\n",
    "\n",
    "        plt.vlines(quants, ymin=0, ymax=np.sum(np.abs(boot_frame[0])), linestyle='--')\n",
    "        plt.xlabel('boot_data')\n",
    "        plt.ylabel('frequency')\n",
    "        plt.title(\"Histogram of boot_data\")\n",
    "        plt.show()\n",
    "       \n",
    "    return {\"boot_data\": boot_data, \n",
    "            \"quants\": quants, \n",
    "            \"p_value\": p_value}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "fe737142-ce36-4093-b3b8-39babcd46822",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.888879150351448"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "N = 1000\n",
    "sample_1 = np.random.normal(0.5, 1, N)\n",
    "sample_2 = np.random.normal(0.5, 1, N)\n",
    "\n",
    "res = get_poisson_bootstrap(sample_1, sample_2, B = 10000, vis=True)\n",
    "res[\"p_value\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "id": "ad87752e-2b62-4994-ba37-c3df2816b62c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.005037865527425217"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "N = 1000\n",
    "sample_1 = np.random.normal(0.5, 1, N)\n",
    "sample_2 = np.random.normal(0.5, 1, N) + 0.1\n",
    "\n",
    "res = get_poisson_bootstrap(sample_1, sample_2, B = 10000, vis=True)\n",
    "res[\"p_value\"]"
   ]
  }
 ],
 "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
}