{ "cells": [ { "cell_type": "markdown", "id": "e9bf0146", "metadata": {}, "source": [ "# 13 - Modelos com Memória" ] }, { "cell_type": "markdown", "id": "16ed1f97", "metadata": {}, "source": [ "## Imports, loadings and functions" ] }, { "cell_type": "code", "execution_count": 1, "id": "39e7ebea", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "\n", "from scipy import stats\n", "\n", "import matplotlib.pyplot as plt\n", "from matplotlib.gridspec import GridSpec\n", "\n", "import pandas as pd\n", "\n", "import networkx as nx\n", "# from causalgraphicalmodels import CausalGraphicalModel\n", "\n", "import arviz as az\n", "# ArviZ ships with style sheets!\n", "# https://python.arviz.org/en/stable/examples/styles.html#example-styles\n", "az.style.use(\"arviz-darkgrid\")\n", "\n", "import xarray as xr\n", "\n", "import stan\n", "import nest_asyncio\n", "\n", "plt.style.use('default')\n", "plt.rcParams['axes.facecolor'] = 'lightgray'\n", "\n", "# To DAG's\n", "import daft\n", "# from causalgraphicalmodels import CausalGraphicalModel # Just work in < python3.9 " ] }, { "cell_type": "code", "execution_count": 2, "id": "7a78c92f", "metadata": {}, "outputs": [], "source": [ "# Add fonts to matplotlib to run xkcd\n", "\n", "from matplotlib import font_manager\n", "\n", "font_dirs = [\"fonts/\"] # The path to the custom font file.\n", "font_files = font_manager.findSystemFonts(fontpaths=font_dirs)\n", "\n", "for font_file in font_files:\n", " font_manager.fontManager.addfont(font_file)" ] }, { "cell_type": "code", "execution_count": 3, "id": "57b999a1", "metadata": {}, "outputs": [], "source": [ "# To make plots like drawing \n", "# plt.xkcd()" ] }, { "cell_type": "code", "execution_count": 4, "id": "8781d488", "metadata": {}, "outputs": [], "source": [ "# To running the stan in jupyter notebook\n", "nest_asyncio.apply()" ] }, { "cell_type": "code", "execution_count": 5, "id": "bbe1f6a8", "metadata": {}, "outputs": [], "source": [ "# Utils functions\n", "\n", "def logit(p):\n", " return np.log(p) - np.log(1 - p)\n", "\n", "def inv_logit(p):\n", " return np.exp(p) / (1 + np.exp(p))" ] }, { "cell_type": "markdown", "id": "ab28e2ba", "metadata": {}, "source": [ "## 13.1 Example: Multilevel tadpoles" ] }, { "cell_type": "markdown", "id": "e0f4702d", "metadata": {}, "source": [ "### R Code 13.1" ] }, { "cell_type": "code", "execution_count": 6, "id": "2966b0de", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
densitypredsizesurvpropsurvtank
010nobig90.9000001
110nobig101.0000002
210nobig70.7000003
310nobig101.0000004
410nosmall90.9000005
510nosmall90.9000006
610nosmall101.0000007
710nosmall90.9000008
810predbig40.4000009
910predbig90.90000010
1010predbig70.70000011
1110predbig60.60000012
1210predsmall70.70000013
1310predsmall50.50000014
1410predsmall90.90000015
1510predsmall90.90000016
1625nobig240.96000017
1725nobig230.92000018
1825nobig220.88000019
1925nobig251.00000020
2025nosmall230.92000021
2125nosmall230.92000022
2225nosmall230.92000023
2325nosmall210.84000024
2425predbig60.24000025
2525predbig130.52000026
2625predbig40.16000027
2725predbig90.36000028
2825predsmall130.52000029
2925predsmall200.80000030
3025predsmall80.32000031
3125predsmall100.40000032
3235nobig340.97142933
3335nobig330.94285734
3435nobig330.94285735
3535nobig310.88571436
3635nosmall310.88571437
3735nosmall351.00000038
3835nosmall330.94285739
3935nosmall320.91428640
4035predbig40.11428641
4135predbig120.34285742
4235predbig130.37142943
4335predbig140.40000044
4435predsmall220.62857145
4535predsmall120.34285746
4635predsmall310.88571447
4735predsmall170.48571448
\n", "
" ], "text/plain": [ " density pred size surv propsurv tank\n", "0 10 no big 9 0.900000 1\n", "1 10 no big 10 1.000000 2\n", "2 10 no big 7 0.700000 3\n", "3 10 no big 10 1.000000 4\n", "4 10 no small 9 0.900000 5\n", "5 10 no small 9 0.900000 6\n", "6 10 no small 10 1.000000 7\n", "7 10 no small 9 0.900000 8\n", "8 10 pred big 4 0.400000 9\n", "9 10 pred big 9 0.900000 10\n", "10 10 pred big 7 0.700000 11\n", "11 10 pred big 6 0.600000 12\n", "12 10 pred small 7 0.700000 13\n", "13 10 pred small 5 0.500000 14\n", "14 10 pred small 9 0.900000 15\n", "15 10 pred small 9 0.900000 16\n", "16 25 no big 24 0.960000 17\n", "17 25 no big 23 0.920000 18\n", "18 25 no big 22 0.880000 19\n", "19 25 no big 25 1.000000 20\n", "20 25 no small 23 0.920000 21\n", "21 25 no small 23 0.920000 22\n", "22 25 no small 23 0.920000 23\n", "23 25 no small 21 0.840000 24\n", "24 25 pred big 6 0.240000 25\n", "25 25 pred big 13 0.520000 26\n", "26 25 pred big 4 0.160000 27\n", "27 25 pred big 9 0.360000 28\n", "28 25 pred small 13 0.520000 29\n", "29 25 pred small 20 0.800000 30\n", "30 25 pred small 8 0.320000 31\n", "31 25 pred small 10 0.400000 32\n", "32 35 no big 34 0.971429 33\n", "33 35 no big 33 0.942857 34\n", "34 35 no big 33 0.942857 35\n", "35 35 no big 31 0.885714 36\n", "36 35 no small 31 0.885714 37\n", "37 35 no small 35 1.000000 38\n", "38 35 no small 33 0.942857 39\n", "39 35 no small 32 0.914286 40\n", "40 35 pred big 4 0.114286 41\n", "41 35 pred big 12 0.342857 42\n", "42 35 pred big 13 0.371429 43\n", "43 35 pred big 14 0.400000 44\n", "44 35 pred small 22 0.628571 45\n", "45 35 pred small 12 0.342857 46\n", "46 35 pred small 31 0.885714 47\n", "47 35 pred small 17 0.485714 48" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv('./data/reedfrogs.csv', sep=\";\")\n", "df['tank'] = df.index.to_list()\n", "df['tank'] += 1 # index start from 1 like Stan works\n", "df" ] }, { "cell_type": "code", "execution_count": 7, "id": "a877a7e4", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
densitysurvpropsurvtank
count48.00000048.00000048.00000048.00
mean23.33333316.3125000.72160724.50
std10.3827469.8847750.26641614.00
min10.0000004.0000000.1142861.00
25%10.0000009.0000000.49642912.75
50%25.00000012.5000000.88571424.50
75%35.00000023.0000000.92000036.25
max35.00000035.0000001.00000048.00
\n", "
" ], "text/plain": [ " density surv propsurv tank\n", "count 48.000000 48.000000 48.000000 48.00\n", "mean 23.333333 16.312500 0.721607 24.50\n", "std 10.382746 9.884775 0.266416 14.00\n", "min 10.000000 4.000000 0.114286 1.00\n", "25% 10.000000 9.000000 0.496429 12.75\n", "50% 25.000000 12.500000 0.885714 24.50\n", "75% 35.000000 23.000000 0.920000 36.25\n", "max 35.000000 35.000000 1.000000 48.00" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.describe()" ] }, { "cell_type": "markdown", "id": "07571408", "metadata": {}, "source": [ "### R Code 13.2" ] }, { "cell_type": "markdown", "id": "13877b50", "metadata": {}, "source": [ "$$ S_i \\sim Binomial(N_i, p_i) $$\n", "\n", "$$ logit(p_i) = \\alpha_{TANK[i]} $$\n", "\n", "$$ \\alpha_j \\sim Normal(0, 1.5), \\mbox{ for } j \\in \\{1, 48\\}$$" ] }, { "cell_type": "code", "execution_count": 8, "id": "6f4c2bd2", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 50% (4000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 75% (6000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 2.3e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.23 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 2.4e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.24 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 4.5e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.45 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 4.4e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.44 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int qty;\n", " array[qty] int N; // Total quantities that have tadpoles in tank\n", " array[qty] int survival; // How many tadpoles survival\n", " array[qty] int tank; // Tank index\n", " }\n", " \n", " parameters {\n", " vector[qty] alpha;\n", " }\n", " \n", " model {\n", " vector[qty] p;\n", " \n", " alpha ~ normal(0, 1.5);\n", " \n", " for (i in 1:qty){\n", " p[i] = alpha[ tank[i] ];\n", " p[i] = inv_logit(alpha[i]);\n", " }\n", " \n", " survival ~ binomial(N, p);\n", " \n", " }\n", "\"\"\"\n", "\n", "dat_list = {\n", " 'qty': len(df),\n", " 'tank': df['tank'].to_list(),\n", " 'survival': df['surv'].to_list(),\n", " 'N': df['density'].to_list()\n", "}\n", "\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 9, "id": "b4d68a9d", "metadata": {}, "outputs": [], "source": [ "model_13_1 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list.keys()\n", ")" ] }, { "cell_type": "code", "execution_count": 10, "id": "5a9709aa", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
alpha[0]1.6880.7690.4922.8930.0090.0087368.03077.01.0
alpha[1]2.4130.9021.0623.8790.0120.0106390.02800.01.0
alpha[2]0.7540.634-0.2871.7100.0080.0086016.02649.01.0
alpha[3]2.4160.9240.9413.7810.0120.0107037.02718.01.0
alpha[4]1.7230.7620.4802.9290.0110.0095394.03054.01.0
alpha[5]1.7080.7640.4572.8600.0100.0095989.02731.01.0
alpha[6]2.3720.8710.9723.7540.0110.0096244.03083.01.0
alpha[7]1.7140.7830.4902.9320.0110.0105967.02391.01.0
alpha[8]-0.3580.613-1.3360.5850.0070.0098826.02906.01.0
alpha[9]1.6990.7600.5302.9500.0100.0096172.02938.01.0
alpha[10]0.7670.672-0.2871.8370.0090.0095800.02373.01.0
alpha[11]0.3870.619-0.5461.4370.0080.0106318.02991.01.0
alpha[12]0.7530.633-0.2271.7600.0080.0085907.02663.01.0
alpha[13]0.0110.593-0.9200.9700.0070.0106565.03111.01.0
alpha[14]1.6950.7380.4242.7660.0100.0086344.03033.01.0
alpha[15]1.6970.7660.4912.9180.0100.0095959.02758.01.0
alpha[16]2.5350.6701.4513.5530.0090.0075782.02985.01.0
alpha[17]2.1500.6211.1553.1350.0080.0076126.02599.01.0
alpha[18]1.8030.5440.9502.6680.0070.0056310.02825.01.0
alpha[19]3.0990.8161.7664.2810.0110.0096216.02690.01.0
alpha[20]2.1320.6111.1012.9980.0090.0075912.02167.01.0
alpha[21]2.1320.5861.2553.0730.0080.0066331.02937.01.0
alpha[22]2.1420.6091.1493.0640.0080.0076008.02591.01.0
alpha[23]1.5420.4940.7162.2500.0060.0057678.02916.01.0
alpha[24]-1.0930.440-1.771-0.3930.0060.0056132.02669.01.0
alpha[25]0.0720.402-0.5240.7500.0050.0076809.03013.01.0
alpha[26]-1.5310.500-2.299-0.7380.0060.0056115.02686.01.0
alpha[27]-0.5560.407-1.2510.0330.0050.0056790.02361.01.0
alpha[28]0.0750.391-0.5360.6750.0050.0066386.03011.01.0
alpha[29]1.3100.4630.5712.0410.0060.0055811.02924.01.0
alpha[30]-0.7290.417-1.361-0.0320.0050.0056228.02635.01.0
alpha[31]-0.3950.404-1.0500.2280.0050.0055725.03194.01.0
alpha[32]2.8440.6561.7373.7950.0080.0076781.02880.01.0
alpha[33]2.4620.5601.5373.3140.0070.0067102.02540.01.0
alpha[34]2.4630.5721.4853.2660.0080.0065946.02926.01.0
alpha[35]1.9000.4801.1622.6730.0070.0055548.03011.01.0
alpha[36]1.9140.4911.1662.6950.0070.0056016.02410.01.0
alpha[37]3.3660.7882.0194.4790.0110.0086427.02749.01.0
alpha[38]2.4510.5821.5443.3340.0070.0067286.02587.01.0
alpha[39]2.1770.5421.3203.0090.0080.0065378.02796.01.0
alpha[40]-1.9070.470-2.663-1.1880.0060.0057052.03063.01.0
alpha[41]-0.6410.358-1.203-0.0660.0040.0046959.02868.01.0
alpha[42]-0.5050.343-1.0780.0260.0040.0045933.03170.01.0
alpha[43]-0.3880.334-0.8840.1740.0040.0046316.03112.01.0
alpha[44]0.5090.345-0.0711.0270.0040.0046490.02912.01.0
alpha[45]-0.6310.344-1.186-0.1000.0040.0046241.02375.01.0
alpha[46]1.9110.4871.0932.6260.0060.0055920.02817.01.0
alpha[47]-0.0610.339-0.5990.4840.0040.0056429.02808.01.0
\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "alpha[0] 1.688 0.769 0.492 2.893 0.009 0.008 7368.0 \n", "alpha[1] 2.413 0.902 1.062 3.879 0.012 0.010 6390.0 \n", "alpha[2] 0.754 0.634 -0.287 1.710 0.008 0.008 6016.0 \n", "alpha[3] 2.416 0.924 0.941 3.781 0.012 0.010 7037.0 \n", "alpha[4] 1.723 0.762 0.480 2.929 0.011 0.009 5394.0 \n", "alpha[5] 1.708 0.764 0.457 2.860 0.010 0.009 5989.0 \n", "alpha[6] 2.372 0.871 0.972 3.754 0.011 0.009 6244.0 \n", "alpha[7] 1.714 0.783 0.490 2.932 0.011 0.010 5967.0 \n", "alpha[8] -0.358 0.613 -1.336 0.585 0.007 0.009 8826.0 \n", "alpha[9] 1.699 0.760 0.530 2.950 0.010 0.009 6172.0 \n", "alpha[10] 0.767 0.672 -0.287 1.837 0.009 0.009 5800.0 \n", "alpha[11] 0.387 0.619 -0.546 1.437 0.008 0.010 6318.0 \n", "alpha[12] 0.753 0.633 -0.227 1.760 0.008 0.008 5907.0 \n", "alpha[13] 0.011 0.593 -0.920 0.970 0.007 0.010 6565.0 \n", "alpha[14] 1.695 0.738 0.424 2.766 0.010 0.008 6344.0 \n", "alpha[15] 1.697 0.766 0.491 2.918 0.010 0.009 5959.0 \n", "alpha[16] 2.535 0.670 1.451 3.553 0.009 0.007 5782.0 \n", "alpha[17] 2.150 0.621 1.155 3.135 0.008 0.007 6126.0 \n", "alpha[18] 1.803 0.544 0.950 2.668 0.007 0.005 6310.0 \n", "alpha[19] 3.099 0.816 1.766 4.281 0.011 0.009 6216.0 \n", "alpha[20] 2.132 0.611 1.101 2.998 0.009 0.007 5912.0 \n", "alpha[21] 2.132 0.586 1.255 3.073 0.008 0.006 6331.0 \n", "alpha[22] 2.142 0.609 1.149 3.064 0.008 0.007 6008.0 \n", "alpha[23] 1.542 0.494 0.716 2.250 0.006 0.005 7678.0 \n", "alpha[24] -1.093 0.440 -1.771 -0.393 0.006 0.005 6132.0 \n", "alpha[25] 0.072 0.402 -0.524 0.750 0.005 0.007 6809.0 \n", "alpha[26] -1.531 0.500 -2.299 -0.738 0.006 0.005 6115.0 \n", "alpha[27] -0.556 0.407 -1.251 0.033 0.005 0.005 6790.0 \n", "alpha[28] 0.075 0.391 -0.536 0.675 0.005 0.006 6386.0 \n", "alpha[29] 1.310 0.463 0.571 2.041 0.006 0.005 5811.0 \n", "alpha[30] -0.729 0.417 -1.361 -0.032 0.005 0.005 6228.0 \n", "alpha[31] -0.395 0.404 -1.050 0.228 0.005 0.005 5725.0 \n", "alpha[32] 2.844 0.656 1.737 3.795 0.008 0.007 6781.0 \n", "alpha[33] 2.462 0.560 1.537 3.314 0.007 0.006 7102.0 \n", "alpha[34] 2.463 0.572 1.485 3.266 0.008 0.006 5946.0 \n", "alpha[35] 1.900 0.480 1.162 2.673 0.007 0.005 5548.0 \n", "alpha[36] 1.914 0.491 1.166 2.695 0.007 0.005 6016.0 \n", "alpha[37] 3.366 0.788 2.019 4.479 0.011 0.008 6427.0 \n", "alpha[38] 2.451 0.582 1.544 3.334 0.007 0.006 7286.0 \n", "alpha[39] 2.177 0.542 1.320 3.009 0.008 0.006 5378.0 \n", "alpha[40] -1.907 0.470 -2.663 -1.188 0.006 0.005 7052.0 \n", "alpha[41] -0.641 0.358 -1.203 -0.066 0.004 0.004 6959.0 \n", "alpha[42] -0.505 0.343 -1.078 0.026 0.004 0.004 5933.0 \n", "alpha[43] -0.388 0.334 -0.884 0.174 0.004 0.004 6316.0 \n", "alpha[44] 0.509 0.345 -0.071 1.027 0.004 0.004 6490.0 \n", "alpha[45] -0.631 0.344 -1.186 -0.100 0.004 0.004 6241.0 \n", "alpha[46] 1.911 0.487 1.093 2.626 0.006 0.005 5920.0 \n", "alpha[47] -0.061 0.339 -0.599 0.484 0.004 0.005 6429.0 \n", "\n", " ess_tail r_hat \n", "alpha[0] 3077.0 1.0 \n", "alpha[1] 2800.0 1.0 \n", "alpha[2] 2649.0 1.0 \n", "alpha[3] 2718.0 1.0 \n", "alpha[4] 3054.0 1.0 \n", "alpha[5] 2731.0 1.0 \n", "alpha[6] 3083.0 1.0 \n", "alpha[7] 2391.0 1.0 \n", "alpha[8] 2906.0 1.0 \n", "alpha[9] 2938.0 1.0 \n", "alpha[10] 2373.0 1.0 \n", "alpha[11] 2991.0 1.0 \n", "alpha[12] 2663.0 1.0 \n", "alpha[13] 3111.0 1.0 \n", "alpha[14] 3033.0 1.0 \n", "alpha[15] 2758.0 1.0 \n", "alpha[16] 2985.0 1.0 \n", "alpha[17] 2599.0 1.0 \n", "alpha[18] 2825.0 1.0 \n", "alpha[19] 2690.0 1.0 \n", "alpha[20] 2167.0 1.0 \n", "alpha[21] 2937.0 1.0 \n", "alpha[22] 2591.0 1.0 \n", "alpha[23] 2916.0 1.0 \n", "alpha[24] 2669.0 1.0 \n", "alpha[25] 3013.0 1.0 \n", "alpha[26] 2686.0 1.0 \n", "alpha[27] 2361.0 1.0 \n", "alpha[28] 3011.0 1.0 \n", "alpha[29] 2924.0 1.0 \n", "alpha[30] 2635.0 1.0 \n", "alpha[31] 3194.0 1.0 \n", "alpha[32] 2880.0 1.0 \n", "alpha[33] 2540.0 1.0 \n", "alpha[34] 2926.0 1.0 \n", "alpha[35] 3011.0 1.0 \n", "alpha[36] 2410.0 1.0 \n", "alpha[37] 2749.0 1.0 \n", "alpha[38] 2587.0 1.0 \n", "alpha[39] 2796.0 1.0 \n", "alpha[40] 3063.0 1.0 \n", "alpha[41] 2868.0 1.0 \n", "alpha[42] 3170.0 1.0 \n", "alpha[43] 3112.0 1.0 \n", "alpha[44] 2912.0 1.0 \n", "alpha[45] 2375.0 1.0 \n", "alpha[46] 2817.0 1.0 \n", "alpha[47] 2808.0 1.0 " ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_1, hdi_prob=0.89)" ] }, { "cell_type": "code", "execution_count": 11, "id": "fe0fa372", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "az.plot_forest(model_13_1, hdi_prob=0.89, combined=True, figsize=(17, 20))\n", "\n", "plt.grid(axis='y', c='white', alpha=0.3)\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "d19e5e79", "metadata": {}, "source": [ "### R Code 13.3" ] }, { "cell_type": "markdown", "id": "a3b894a1", "metadata": {}, "source": [ "#### Multilevel model Tadpole\n", "\n", "$$ S_i \\sim Binomial(N_i, p_i) $$\n", "\n", "$$ logit(p_i) = \\alpha_{TANK[i]} $$\n", "\n", "$$ \\alpha[j] \\sim Normal(\\bar{\\alpha}, \\sigma) \\mbox{ - [Adaptative prior]} $$\n", "\n", "$$ \\bar{\\alpha} \\sim Normal(0, 1.5) \\mbox{ - [prior to average tank]} $$\n", "\n", "$$ \\sigma \\sim Exponential(1) \\mbox{ - [prior for standard deviation of tanks]} $$" ] }, { "cell_type": "code", "execution_count": 12, "id": "7203c437", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 50% (4000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 75% (6000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 3.1e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.31 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 2.4e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.24 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 3e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.3 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_e2h0pxzb/model_jyvtw2so.stan', line 18, column 8 to column 41)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 3.7e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.37 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int qty;\n", " array[qty] int N;\n", " array[qty] int survival;\n", " array[qty] int tank;\n", " }\n", " \n", " parameters {\n", " vector[qty] alpha;\n", " real bar_alpha;\n", " real sigma;\n", " }\n", " \n", " model {\n", " vector[qty] p;\n", " \n", " alpha ~ normal(bar_alpha, sigma);\n", " \n", " bar_alpha ~ normal(0, 1.5);\n", " sigma ~ exponential(1);\n", " \n", " for (i in 1:qty){\n", " p[i] = alpha[ tank[i] ];\n", " p[i] = inv_logit(p[i]);\n", " }\n", " \n", " survival ~ binomial(N, p);\n", " }\n", "\"\"\"\n", "\n", "\n", "dat_list = {\n", " 'qty': len(df),\n", " 'tank': df['tank'].to_list(),\n", " 'survival': df['surv'].to_list(),\n", " 'N': df['density'].to_list()\n", "}\n", "\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 13, "id": "e312974d", "metadata": {}, "outputs": [], "source": [ "model_13_2 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list.keys(),\n", ")" ] }, { "cell_type": "code", "execution_count": 14, "id": "2732fcfd", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
alpha[0]2.1220.8670.7613.4100.0130.0114670.02652.01.0
alpha[1]3.0571.1111.3314.7800.0170.0144514.02725.01.0
alpha[2]1.0090.668-0.0922.0140.0100.0094757.02303.01.0
alpha[3]3.0511.1341.2874.8170.0190.0153884.02546.01.0
alpha[4]2.1130.8630.7753.5120.0130.0114661.02460.01.0
alpha[5]2.1240.8670.6963.4130.0150.0123970.02189.01.0
alpha[6]3.0661.1251.2384.7470.0190.0154090.02394.01.0
alpha[7]2.1160.8530.7513.4190.0120.0105383.02966.01.0
alpha[8]-0.1680.633-1.1620.8470.0090.0104716.02678.01.0
alpha[9]2.1100.8460.7433.4030.0120.0105010.02650.01.0
alpha[10]0.9870.667-0.0892.0130.0100.0084779.02951.01.0
alpha[11]0.5970.646-0.4111.6040.0090.0094973.02600.01.0
alpha[12]1.0020.671-0.0552.0350.0090.0085171.02865.01.0
alpha[13]0.2140.609-0.7171.2320.0090.0104279.02814.01.0
alpha[14]2.1230.8590.7703.4710.0130.0104826.02780.01.0
alpha[15]2.1330.8740.6773.4030.0130.0114744.02533.01.0
alpha[16]2.9100.8051.5964.1350.0120.0105070.02605.01.0
alpha[17]2.3840.6521.3583.3930.0100.0084563.02810.01.0
alpha[18]2.0130.5841.0922.9230.0090.0074620.02575.01.0
alpha[19]3.6330.9822.0815.1180.0170.0133893.02324.01.0
alpha[20]2.3860.6541.3163.3780.0100.0074954.02849.01.0
alpha[21]2.3960.6741.3173.4410.0090.0075553.02677.01.0
alpha[22]2.3760.6581.3203.3750.0100.0084411.02340.01.0
alpha[23]1.7020.5220.8672.5030.0080.0064477.02490.01.0
alpha[24]-0.9910.438-1.671-0.2920.0070.0054514.03235.01.0
alpha[25]0.1650.406-0.4760.8100.0060.0064809.03083.01.0
alpha[26]-1.4350.507-2.178-0.5820.0080.0064671.02617.01.0
alpha[27]-0.4760.406-1.1290.1610.0060.0054928.03110.01.0
alpha[28]0.1550.392-0.5050.7400.0050.0075297.02213.01.0
alpha[29]1.4400.4660.6892.1820.0060.0055406.03145.01.0
alpha[30]-0.6320.405-1.274-0.0050.0060.0055142.03273.01.0
alpha[31]-0.3090.394-0.9130.3400.0050.0055375.03214.01.0
alpha[32]3.1880.7721.9804.3770.0130.0104251.02390.01.0
alpha[33]2.6960.6371.7093.6760.0100.0084229.02695.01.0
alpha[34]2.7220.6451.7173.7330.0100.0084610.02941.01.0
alpha[35]2.0490.5011.2812.8320.0070.0064958.02795.01.0
alpha[36]2.0670.5041.3222.9100.0080.0064683.02858.01.0
alpha[37]3.8620.9632.3665.3310.0160.0124501.02430.01.0
alpha[38]2.7040.6521.6033.6770.0100.0084824.02751.01.0
alpha[39]2.3520.5751.4493.2430.0090.0074457.02668.01.0
alpha[40]-1.8080.475-2.564-1.0830.0070.0054606.02552.01.0
alpha[41]-0.5740.349-1.119-0.0200.0040.0046263.03038.01.0
alpha[42]-0.4520.346-1.0060.0940.0050.0044843.02948.01.0
alpha[43]-0.3380.336-0.8450.2270.0050.0044978.02791.01.0
alpha[44]0.5790.3500.0381.1450.0050.0044657.02913.01.0
alpha[45]-0.5730.347-1.1060.0030.0050.0045124.02977.01.0
alpha[46]2.0830.5241.2822.9380.0080.0064687.02834.01.0
alpha[47]-0.0010.323-0.5160.5110.0050.0054878.03390.01.0
bar_alpha1.3380.2520.9421.7400.0040.0033387.02606.01.0
sigma1.6130.2141.2741.9290.0040.0032393.02698.01.0
\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "alpha[0] 2.122 0.867 0.761 3.410 0.013 0.011 4670.0 \n", "alpha[1] 3.057 1.111 1.331 4.780 0.017 0.014 4514.0 \n", "alpha[2] 1.009 0.668 -0.092 2.014 0.010 0.009 4757.0 \n", "alpha[3] 3.051 1.134 1.287 4.817 0.019 0.015 3884.0 \n", "alpha[4] 2.113 0.863 0.775 3.512 0.013 0.011 4661.0 \n", "alpha[5] 2.124 0.867 0.696 3.413 0.015 0.012 3970.0 \n", "alpha[6] 3.066 1.125 1.238 4.747 0.019 0.015 4090.0 \n", "alpha[7] 2.116 0.853 0.751 3.419 0.012 0.010 5383.0 \n", "alpha[8] -0.168 0.633 -1.162 0.847 0.009 0.010 4716.0 \n", "alpha[9] 2.110 0.846 0.743 3.403 0.012 0.010 5010.0 \n", "alpha[10] 0.987 0.667 -0.089 2.013 0.010 0.008 4779.0 \n", "alpha[11] 0.597 0.646 -0.411 1.604 0.009 0.009 4973.0 \n", "alpha[12] 1.002 0.671 -0.055 2.035 0.009 0.008 5171.0 \n", "alpha[13] 0.214 0.609 -0.717 1.232 0.009 0.010 4279.0 \n", "alpha[14] 2.123 0.859 0.770 3.471 0.013 0.010 4826.0 \n", "alpha[15] 2.133 0.874 0.677 3.403 0.013 0.011 4744.0 \n", "alpha[16] 2.910 0.805 1.596 4.135 0.012 0.010 5070.0 \n", "alpha[17] 2.384 0.652 1.358 3.393 0.010 0.008 4563.0 \n", "alpha[18] 2.013 0.584 1.092 2.923 0.009 0.007 4620.0 \n", "alpha[19] 3.633 0.982 2.081 5.118 0.017 0.013 3893.0 \n", "alpha[20] 2.386 0.654 1.316 3.378 0.010 0.007 4954.0 \n", "alpha[21] 2.396 0.674 1.317 3.441 0.009 0.007 5553.0 \n", "alpha[22] 2.376 0.658 1.320 3.375 0.010 0.008 4411.0 \n", "alpha[23] 1.702 0.522 0.867 2.503 0.008 0.006 4477.0 \n", "alpha[24] -0.991 0.438 -1.671 -0.292 0.007 0.005 4514.0 \n", "alpha[25] 0.165 0.406 -0.476 0.810 0.006 0.006 4809.0 \n", "alpha[26] -1.435 0.507 -2.178 -0.582 0.008 0.006 4671.0 \n", "alpha[27] -0.476 0.406 -1.129 0.161 0.006 0.005 4928.0 \n", "alpha[28] 0.155 0.392 -0.505 0.740 0.005 0.007 5297.0 \n", "alpha[29] 1.440 0.466 0.689 2.182 0.006 0.005 5406.0 \n", "alpha[30] -0.632 0.405 -1.274 -0.005 0.006 0.005 5142.0 \n", "alpha[31] -0.309 0.394 -0.913 0.340 0.005 0.005 5375.0 \n", "alpha[32] 3.188 0.772 1.980 4.377 0.013 0.010 4251.0 \n", "alpha[33] 2.696 0.637 1.709 3.676 0.010 0.008 4229.0 \n", "alpha[34] 2.722 0.645 1.717 3.733 0.010 0.008 4610.0 \n", "alpha[35] 2.049 0.501 1.281 2.832 0.007 0.006 4958.0 \n", "alpha[36] 2.067 0.504 1.322 2.910 0.008 0.006 4683.0 \n", "alpha[37] 3.862 0.963 2.366 5.331 0.016 0.012 4501.0 \n", "alpha[38] 2.704 0.652 1.603 3.677 0.010 0.008 4824.0 \n", "alpha[39] 2.352 0.575 1.449 3.243 0.009 0.007 4457.0 \n", "alpha[40] -1.808 0.475 -2.564 -1.083 0.007 0.005 4606.0 \n", "alpha[41] -0.574 0.349 -1.119 -0.020 0.004 0.004 6263.0 \n", "alpha[42] -0.452 0.346 -1.006 0.094 0.005 0.004 4843.0 \n", "alpha[43] -0.338 0.336 -0.845 0.227 0.005 0.004 4978.0 \n", "alpha[44] 0.579 0.350 0.038 1.145 0.005 0.004 4657.0 \n", "alpha[45] -0.573 0.347 -1.106 0.003 0.005 0.004 5124.0 \n", "alpha[46] 2.083 0.524 1.282 2.938 0.008 0.006 4687.0 \n", "alpha[47] -0.001 0.323 -0.516 0.511 0.005 0.005 4878.0 \n", "bar_alpha 1.338 0.252 0.942 1.740 0.004 0.003 3387.0 \n", "sigma 1.613 0.214 1.274 1.929 0.004 0.003 2393.0 \n", "\n", " ess_tail r_hat \n", "alpha[0] 2652.0 1.0 \n", "alpha[1] 2725.0 1.0 \n", "alpha[2] 2303.0 1.0 \n", "alpha[3] 2546.0 1.0 \n", "alpha[4] 2460.0 1.0 \n", "alpha[5] 2189.0 1.0 \n", "alpha[6] 2394.0 1.0 \n", "alpha[7] 2966.0 1.0 \n", "alpha[8] 2678.0 1.0 \n", "alpha[9] 2650.0 1.0 \n", "alpha[10] 2951.0 1.0 \n", "alpha[11] 2600.0 1.0 \n", "alpha[12] 2865.0 1.0 \n", "alpha[13] 2814.0 1.0 \n", "alpha[14] 2780.0 1.0 \n", "alpha[15] 2533.0 1.0 \n", "alpha[16] 2605.0 1.0 \n", "alpha[17] 2810.0 1.0 \n", "alpha[18] 2575.0 1.0 \n", "alpha[19] 2324.0 1.0 \n", "alpha[20] 2849.0 1.0 \n", "alpha[21] 2677.0 1.0 \n", "alpha[22] 2340.0 1.0 \n", "alpha[23] 2490.0 1.0 \n", "alpha[24] 3235.0 1.0 \n", "alpha[25] 3083.0 1.0 \n", "alpha[26] 2617.0 1.0 \n", "alpha[27] 3110.0 1.0 \n", "alpha[28] 2213.0 1.0 \n", "alpha[29] 3145.0 1.0 \n", "alpha[30] 3273.0 1.0 \n", "alpha[31] 3214.0 1.0 \n", "alpha[32] 2390.0 1.0 \n", "alpha[33] 2695.0 1.0 \n", "alpha[34] 2941.0 1.0 \n", "alpha[35] 2795.0 1.0 \n", "alpha[36] 2858.0 1.0 \n", "alpha[37] 2430.0 1.0 \n", "alpha[38] 2751.0 1.0 \n", "alpha[39] 2668.0 1.0 \n", "alpha[40] 2552.0 1.0 \n", "alpha[41] 3038.0 1.0 \n", "alpha[42] 2948.0 1.0 \n", "alpha[43] 2791.0 1.0 \n", "alpha[44] 2913.0 1.0 \n", "alpha[45] 2977.0 1.0 \n", "alpha[46] 2834.0 1.0 \n", "alpha[47] 3390.0 1.0 \n", "bar_alpha 2606.0 1.0 \n", "sigma 2698.0 1.0 " ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_2, hdi_prob=0.89)" ] }, { "cell_type": "code", "execution_count": 15, "id": "29c63ab3", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "az.plot_forest(model_13_2, hdi_prob=0.89, combined=True, figsize=(17, 20))\n", "\n", "plt.grid(axis='y', color='white', alpha=0.3)\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "c81d27cd", "metadata": {}, "source": [ "### R Code 13.4" ] }, { "cell_type": "code", "execution_count": 16, "id": "b2f99ea1", "metadata": {}, "outputs": [], "source": [ "# az.compare(model_13_1, model_13_2)" ] }, { "cell_type": "markdown", "id": "122be960", "metadata": {}, "source": [ "### R Code 13.5" ] }, { "cell_type": "code", "execution_count": 17, "id": "fe08dff6", "metadata": {}, "outputs": [], "source": [ "means = [ model_13_2.posterior.alpha.sel(alpha_dim_0=(i-1)).values.flatten().mean() for i in df.tank ]\n", "means = inv_logit(means)" ] }, { "cell_type": "code", "execution_count": 18, "id": "32f09207", "metadata": {}, "outputs": [], "source": [ "# My test, this is not originaly in book\n", "means_13_1 = [ model_13_1.posterior.alpha.sel(alpha_dim_0=(i-1)).values.flatten().mean() for i in df.tank ]\n", "means_13_1 = inv_logit(means_13_1)" ] }, { "cell_type": "code", "execution_count": 19, "id": "ce91f00b", "metadata": {}, "outputs": [], "source": [ "bar_alpha_log = model_13_2.posterior.bar_alpha.values.flatten()\n", "bar_alpha = inv_logit(bar_alpha_log)\n", "bar_alpha_mean = bar_alpha.mean() \n", "\n", "sigma_log = model_13_2.posterior.sigma.values.flatten()\n", "sigma = inv_logit(sigma_log)\n", "sigma_mean = sigma.mean()" ] }, { "cell_type": "code", "execution_count": 20, "id": "60f9bf96", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(17, 6))\n", "plt.ylim = ([0, 1])\n", "\n", "plt.scatter(df.tank, means, edgecolors='black', c='lightgray', s=100)\n", "plt.scatter(df.tank, means_13_1, edgecolors='yellow', c='lightgray', s=100, alpha=0.4)\n", "plt.scatter(df.tank, df.propsurv, c='blue')\n", "\n", "\n", "plt.axvline(x=15.5, ls='--', color='white', alpha=0.3)\n", "plt.axvline(x=31.5, ls='--', color='white', alpha=0.3)\n", "\n", "plt.axhline(y=bar_alpha_mean, ls='--', c='black', alpha=0.7)\n", "\n", "plt.text(4, 0.05, 'Small Tank', size=12)\n", "plt.text(22, 0.05, 'Medium Tank', size=12)\n", "plt.text(40, 0.05, 'Large Tank', size=12)\n", "\n", "plt.gca().set_ylim(0.0, 1.05)\n", "\n", "plt.title('Tadpole survival Tanks')\n", "plt.xlabel('Tank Index')\n", "plt.ylabel('Porportion Survival')\n", "\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "3e064bb2", "metadata": {}, "source": [ "- **Blue dot**: Proportion survival s_i/N_i\n", "\n", "- **Black circle**: Multilevel model estimative\n", "\n", "- **Light Yellow**: No-pooling estimative" ] }, { "cell_type": "markdown", "id": "1fb23838", "metadata": {}, "source": [ "### R Code 13.6" ] }, { "cell_type": "code", "execution_count": 21, "id": "086fcc9f", "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig = plt.figure(figsize=(17, 6))\n", "gs = GridSpec(1, 2)\n", "\n", "x = np.linspace(-3, 4)\n", "\n", "s_sampled = 500\n", "\n", "ax1 = fig.add_subplot(gs[0, 0])\n", "log_odds_survival = []\n", "log_odds_sampled_index = np.random.choice(len(bar_alpha_log) ,size=s_sampled, replace=False)\n", "\n", "for i in log_odds_sampled_index:\n", " log_odds_survival.append(stats.norm.pdf(x, bar_alpha_log[i], sigma_log[i]))\n", "\n", "for i in range(s_sampled):\n", " ax1.plot(x, log_odds_survival[i], c='darkblue', linewidth=0.05)\n", "ax1.set_title('Survival across Tank')\n", "ax1.set_xlabel('log_odds survival')\n", "ax1.set_ylabel('Density')\n", " \n", "\n", "ax2 = fig.add_subplot(gs[0, 1])\n", "samples_log = np.random.normal(bar_alpha_log, sigma_log)\n", "ax2.hist(inv_logit(samples_log), rwidth=0.9, color='darkblue', density=True)\n", "ax2.axvline(x=np.mean(inv_logit(samples_log)), c='black', ls='--')\n", "ax2.text(np.mean(inv_logit(samples_log))+0.01, 2.5, 'Mean')\n", "\n", "ax2.set_title('Survival probabilities simulations')\n", "ax2.set_xlabel('Probability survival')\n", "ax2.set_ylabel('Density')\n", "\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "81b96139", "metadata": {}, "source": [ "## 13.2 Varing effects and underfitting/overfitting trade-off" ] }, { "cell_type": "markdown", "id": "72b22cfa", "metadata": {}, "source": [ "The model\n", "\n", "$$ S_i \\sim Binomial(N_i, p_i) $$\n", "\n", "$$ logit(p_i) = \\alpha_{POND[i]} $$\n", "\n", "$$ \\alpha_j \\sim Normal(\\bar{\\alpha}, \\sigma) $$\n", "\n", "$$ \\bar{\\alpha} \\sim Normal(0, 1.5) $$\n", "\n", "$$ \\sigma \\sim Exponential(1) $$" ] }, { "cell_type": "markdown", "id": "a6fef872", "metadata": {}, "source": [ "$\\bar{\\alpha} := $ the avegare log-oods fo survival in the entire population of ponds\n", "\n", "$\\sigma := $ the standard deviation of the distribution of log-oods of survivial among ponds\n", "\n", "$\\alpha := $ a vector of individual pond intercepts, one for each pond" ] }, { "cell_type": "markdown", "id": "1ebf32aa", "metadata": {}, "source": [ "### R Code 13.7" ] }, { "cell_type": "code", "execution_count": 22, "id": "4279f35b", "metadata": {}, "outputs": [], "source": [ "a_bar = 1.5\n", "sigma = 1.5\n", "nponds = 60\n", "\n", "repeats = 15\n", "\n", "Ni = np.repeat([5, 10, 25, 35], repeats=repeats)" ] }, { "cell_type": "markdown", "id": "646a770c", "metadata": {}, "source": [ "### R Code 13.8" ] }, { "cell_type": "code", "execution_count": 23, "id": "7d9ecd3d", "metadata": {}, "outputs": [], "source": [ "a_pond = np.random.normal(loc=a_bar, scale=sigma, size=nponds)" ] }, { "cell_type": "markdown", "id": "ef39008e", "metadata": {}, "source": [ "### R Code 13.9" ] }, { "cell_type": "code", "execution_count": 24, "id": "c9614282", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
pondNitrue_a
0152.739710
1253.174035
2351.950330
3450.743275
4550.167714
\n", "
" ], "text/plain": [ " pond Ni true_a\n", "0 1 5 2.739710\n", "1 2 5 3.174035\n", "2 3 5 1.950330\n", "3 4 5 0.743275\n", "4 5 5 0.167714" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "d = {\n", " 'pond': np.arange(nponds) + 1,\n", " 'Ni':Ni,\n", " 'true_a': a_pond,\n", "}\n", "\n", "dsim = pd.DataFrame(data=d)\n", "dsim.head()" ] }, { "cell_type": "markdown", "id": "bdfe8ced", "metadata": {}, "source": [ "### R Code 13.10" ] }, { "cell_type": "code", "execution_count": 25, "id": "6ac31a70", "metadata": {}, "outputs": [], "source": [ "# Code in R -> integer vs numeric" ] }, { "cell_type": "markdown", "id": "f6abedc7", "metadata": {}, "source": [ "### R Code 13.11" ] }, { "cell_type": "code", "execution_count": 26, "id": "624e38f2", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
pondNitrue_aSi
0152.7397105
1253.1740354
2351.9503305
3450.7432752
4550.1677143
\n", "
" ], "text/plain": [ " pond Ni true_a Si\n", "0 1 5 2.739710 5\n", "1 2 5 3.174035 4\n", "2 3 5 1.950330 5\n", "3 4 5 0.743275 2\n", "4 5 5 0.167714 3" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dsim['Si'] = np.random.binomial(n=dsim['Ni'], p=inv_logit(dsim['true_a']))\n", "dsim.head()" ] }, { "cell_type": "markdown", "id": "1039fcb5", "metadata": {}, "source": [ "### R Code 13.12" ] }, { "cell_type": "markdown", "id": "8b394277", "metadata": {}, "source": [ "#### 13.2.4 Compute the no-pooling estimates" ] }, { "cell_type": "code", "execution_count": 27, "id": "27491eda", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
pondNitrue_aSip_nopool
0152.73971051.0
1253.17403540.8
2351.95033051.0
3450.74327520.4
4550.16771430.6
\n", "
" ], "text/plain": [ " pond Ni true_a Si p_nopool\n", "0 1 5 2.739710 5 1.0\n", "1 2 5 3.174035 4 0.8\n", "2 3 5 1.950330 5 1.0\n", "3 4 5 0.743275 2 0.4\n", "4 5 5 0.167714 3 0.6" ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dsim['p_nopool'] = dsim['Si'] / dsim['Ni']\n", "dsim.head()" ] }, { "cell_type": "markdown", "id": "99d028cf", "metadata": {}, "source": [ "### R Code 13.13" ] }, { "cell_type": "markdown", "id": "99ccaee8", "metadata": {}, "source": [ "#### 13.2.5 Compute the partial-pooling estimates" ] }, { "cell_type": "code", "execution_count": 28, "id": "2681a744", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 50% (4000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 75% (6000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 81% (6500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 4.9e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.49 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 4e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.4 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 5.1e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.51 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 8.1e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.81 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int N;\n", " array[N] int pond; // Pond index\n", " array[N] int Ni; // Population in pond[i]\n", " array[N] int Si; // Survivals from Ni pond\n", " }\n", " \n", " parameters {\n", " vector[N] alpha;\n", " real bar_alpha;\n", " real sigma;\n", " }\n", " \n", " model {\n", " vector[N] pi;\n", " \n", " // Link\n", " for (i in 1:N) {\n", " pi[i] = alpha[ pond[i] ];\n", " pi[i] = inv_logit(pi[i]);\n", " }\n", " \n", " // Prior\n", " alpha ~ normal(bar_alpha, sigma);\n", " \n", " // Hyper Prior\n", " bar_alpha ~ normal(0, 1.5);\n", " sigma ~ exponential(1);\n", " \n", " // Likelihood\n", " Si ~ binomial(Ni, pi);\n", " }\n", "\"\"\"\n", "\n", "\n", "dat_list = {\n", " 'N': len(dsim),\n", " 'Ni': dsim['Ni'].to_list(),\n", " 'pond': dsim['pond'].to_list(),\n", " 'Si': dsim['Si'].to_list(),\n", "}\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "markdown", "id": "cb5b2d66", "metadata": {}, "source": [ "### R Code 13.14" ] }, { "cell_type": "code", "execution_count": 29, "id": "7eb8e89c", "metadata": {}, "outputs": [], "source": [ "model_13_3 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list\n", ")" ] }, { "cell_type": "code", "execution_count": 30, "id": "8dd12db9", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
alpha[0]2.8131.2360.7854.6600.0210.0163797.02736.01.0
alpha[1]1.6250.9910.0743.2260.0150.0134780.02616.01.0
alpha[2]2.8161.2720.8624.8080.0200.0174445.02414.01.0
alpha[3]0.0660.851-1.3481.3670.0120.0165374.02548.01.0
alpha[4]0.7430.863-0.5672.1570.0120.0125583.02624.01.0
..............................
alpha[57]3.2690.7962.0304.5310.0120.0104592.02290.01.0
alpha[58]-1.4110.420-2.025-0.7130.0060.0045201.02968.01.0
alpha[59]-0.4390.335-0.9560.0950.0050.0044896.02723.01.0
bar_alpha1.5510.2531.1661.9750.0050.0032812.03216.01.0
sigma1.6570.2301.3002.0190.0060.0041418.01740.01.0
\n", "

62 rows × 9 columns

\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "alpha[0] 2.813 1.236 0.785 4.660 0.021 0.016 3797.0 \n", "alpha[1] 1.625 0.991 0.074 3.226 0.015 0.013 4780.0 \n", "alpha[2] 2.816 1.272 0.862 4.808 0.020 0.017 4445.0 \n", "alpha[3] 0.066 0.851 -1.348 1.367 0.012 0.016 5374.0 \n", "alpha[4] 0.743 0.863 -0.567 2.157 0.012 0.012 5583.0 \n", "... ... ... ... ... ... ... ... \n", "alpha[57] 3.269 0.796 2.030 4.531 0.012 0.010 4592.0 \n", "alpha[58] -1.411 0.420 -2.025 -0.713 0.006 0.004 5201.0 \n", "alpha[59] -0.439 0.335 -0.956 0.095 0.005 0.004 4896.0 \n", "bar_alpha 1.551 0.253 1.166 1.975 0.005 0.003 2812.0 \n", "sigma 1.657 0.230 1.300 2.019 0.006 0.004 1418.0 \n", "\n", " ess_tail r_hat \n", "alpha[0] 2736.0 1.0 \n", "alpha[1] 2616.0 1.0 \n", "alpha[2] 2414.0 1.0 \n", "alpha[3] 2548.0 1.0 \n", "alpha[4] 2624.0 1.0 \n", "... ... ... \n", "alpha[57] 2290.0 1.0 \n", "alpha[58] 2968.0 1.0 \n", "alpha[59] 2723.0 1.0 \n", "bar_alpha 3216.0 1.0 \n", "sigma 1740.0 1.0 \n", "\n", "[62 rows x 9 columns]" ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_3, hdi_prob=0.89)" ] }, { "cell_type": "markdown", "id": "999a0658", "metadata": {}, "source": [ "### R Code 13.15" ] }, { "cell_type": "code", "execution_count": 31, "id": "07042ce0", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
pondNitrue_aSip_nopoolp_partpool
0152.73971051.00.943360
1253.17403540.80.835496
2351.95033051.00.943544
3450.74327520.40.516615
4550.16771430.60.677755
\n", "
" ], "text/plain": [ " pond Ni true_a Si p_nopool p_partpool\n", "0 1 5 2.739710 5 1.0 0.943360\n", "1 2 5 3.174035 4 0.8 0.835496\n", "2 3 5 1.950330 5 1.0 0.943544\n", "3 4 5 0.743275 2 0.4 0.516615\n", "4 5 5 0.167714 3 0.6 0.677755" ] }, "execution_count": 31, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dsim['p_partpool'] = [inv_logit(model_13_3.posterior.alpha.sel(alpha_dim_0=i).values.mean()) for i in range(len(dsim))]\n", "dsim.head()" ] }, { "cell_type": "markdown", "id": "3413ccee", "metadata": {}, "source": [ "### R Code 13.16" ] }, { "cell_type": "code", "execution_count": 32, "id": "9a6fd6bb", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
pondNitrue_aSip_nopoolp_partpoolp_true
0152.73971051.00.9433600.939330
1253.17403540.80.8354960.959845
2351.95033051.00.9435440.875483
3450.74327520.40.5166150.677712
4550.16771430.60.6777550.541830
\n", "
" ], "text/plain": [ " pond Ni true_a Si p_nopool p_partpool p_true\n", "0 1 5 2.739710 5 1.0 0.943360 0.939330\n", "1 2 5 3.174035 4 0.8 0.835496 0.959845\n", "2 3 5 1.950330 5 1.0 0.943544 0.875483\n", "3 4 5 0.743275 2 0.4 0.516615 0.677712\n", "4 5 5 0.167714 3 0.6 0.677755 0.541830" ] }, "execution_count": 32, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dsim['p_true'] = inv_logit(dsim['true_a'])\n", "dsim.head()" ] }, { "cell_type": "markdown", "id": "a8170cd3", "metadata": {}, "source": [ "### R Code 13.17" ] }, { "cell_type": "code", "execution_count": 33, "id": "c78e65a9", "metadata": {}, "outputs": [], "source": [ "no_pool_error = np.abs(dsim['p_nopool'] - dsim['p_true'])\n", "partpool_error = np.abs(dsim['p_partpool'] - dsim['p_true'])" ] }, { "cell_type": "markdown", "id": "340feedc", "metadata": {}, "source": [ "### R Code 13.18" ] }, { "cell_type": "code", "execution_count": 34, "id": "f6734480", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(17, 8))\n", "plt.ylim = ([0, 1])\n", "max_lim_graph = 0.3\n", "\n", "plt.scatter(dsim.pond, partpool_error, edgecolors='black', c='lightgray', s=100)\n", "plt.scatter(dsim.pond, no_pool_error, c='blue')\n", "\n", "qty_unique_ponds = len(dsim['Ni'].unique())\n", "qty_each_ponds = repeats # The number of repetitions for each element in each pond.\n", "\n", "\n", "# Vertical lines\n", "for i in range(qty_unique_ponds):\n", " plt.axvline(x=qty_each_ponds*(i+1) + 0.5, ls='--', color='white', alpha=0.3)\n", " \n", " partpool_error_mean = np.mean(partpool_error[(qty_each_ponds*i):(qty_each_ponds*(i+1))])\n", " no_pool_error_mean = np.mean(no_pool_error[(qty_each_ponds*i):(qty_each_ponds*(i+1))])\n", " \n", " plt.hlines(y=partpool_error_mean, xmin=1+(qty_each_ponds*i), xmax=qty_each_ponds+(qty_each_ponds*i), ls='--', colors='black', alpha=0.7)\n", " plt.hlines(y=no_pool_error_mean, xmin=1+(qty_each_ponds*i), xmax=qty_each_ponds+(qty_each_ponds*i), ls='-', colors='blue', alpha=0.7)\n", "\n", "score_no_pooling = 0\n", "score_partial_pooling = 0\n", " \n", "for i in dsim.pond:\n", " if no_pool_error[i-1] >= partpool_error[i-1]: # partial polling is better\n", " plt.vlines(x=i, ymin=no_pool_error[i-1], ymax=partpool_error[i-1], ls='--', colors='green', alpha=0.3)\n", " score_partial_pooling += no_pool_error[i-1] - partpool_error[i-1] # How partial pooling is better\n", " \n", " else: # no pooling is better\n", " plt.vlines(x=i, ymin=no_pool_error[i-1], ymax=partpool_error[i-1], ls='--', colors='red', alpha=0.3)\n", " score_no_pooling += partpool_error[i-1] - no_pool_error[i-1] # How no pooling is better\n", "\n", "plt.text(7, max_lim_graph, 'Tiny Ponds ($5$)', size=12)\n", "plt.text(21, max_lim_graph, 'Small Ponds ($10$)', size=12)\n", "plt.text(35, max_lim_graph, 'Medium Ponds ($25$)', size=12)\n", "plt.text(50, max_lim_graph, 'Large Ponds ($35$)', size=12)\n", "\n", "plt.text(47, 0.25, f'Partial pooling is better by = {round(score_partial_pooling, 2)}')\n", "plt.text(47, 0.24, f'No pooling is better by = {round(score_no_pooling, 2)}')\n", "plt.text(47, 0.23, f'Partial Polling/No polling = {round((score_partial_pooling/score_no_pooling)*100, 2)}%')\n", "\n", "\n", "plt.gca().set_ylim(-0.01, max_lim_graph + 0.05)\n", "\n", "plt.title('Pond survival error absolute \\n\\n Black dash line = partial pooling \\n Blue line = no pooling')\n", "plt.xlabel('Pond Index')\n", "plt.ylabel('Absolute Error')\n", "\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "3d528d29", "metadata": {}, "source": [ "### R Code 13.20" ] }, { "cell_type": "code", "execution_count": 35, "id": "2d7bc2cf", "metadata": {}, "outputs": [], "source": [ "# Reuse code in using Rethinking packages in R, here is automatically reuse!\n", "# Just re-run from R Code 13.7" ] }, { "cell_type": "markdown", "id": "0de0e34d", "metadata": {}, "source": [ "## 13.3 More than one type of cluster" ] }, { "cell_type": "markdown", "id": "e0133bfc", "metadata": {}, "source": [ "#### Multilevel Chimpanzees\n", "\n", "$$ L_i \\sim Binomial(1, p_i) $$\n", "\n", "$$ logit(p_i) = \\alpha_{ACTOR[i]} + \\gamma_{BLOCK[i]} + \\beta_{TREATMENT[i]} $$\n", "\n", "$$ \\beta_j \\sim Normal(0, 0.5) \\mbox{ , } j \\in \\{1, ... ,4\\} $$\n", "\n", "$$ \\alpha_j \\sim Normal(\\bar{\\alpha}, \\sigma_\\alpha) \\mbox{ , } j \\in \\{1, ... ,7\\} $$\n", "\n", "$$ \\gamma_j \\sim Normal(0, \\sigma_\\gamma) \\mbox{ , } j \\in \\{1, ... ,6\\} $$\n", "\n", "$$ \\bar{\\alpha} \\sim Normal(0, 1.5) $$\n", "\n", "$$ \\sigma_{\\alpha} \\sim Exponential(1) $$\n", "\n", "$$ \\sigma_{\\gamma} \\sim Exponential(1) $$" ] }, { "cell_type": "markdown", "id": "7fadb95d", "metadata": {}, "source": [ "### R Code 13.21" ] }, { "cell_type": "code", "execution_count": 36, "id": "be447731", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
actorrecipientconditionblocktrialprosoc_leftchose_prosocpulled_left
01NaN012010
11NaN014001
21NaN016100
31NaN018010
41NaN0110111
\n", "
" ], "text/plain": [ " actor recipient condition block trial prosoc_left chose_prosoc \\\n", "0 1 NaN 0 1 2 0 1 \n", "1 1 NaN 0 1 4 0 0 \n", "2 1 NaN 0 1 6 1 0 \n", "3 1 NaN 0 1 8 0 1 \n", "4 1 NaN 0 1 10 1 1 \n", "\n", " pulled_left \n", "0 0 \n", "1 1 \n", "2 0 \n", "3 0 \n", "4 1 " ] }, "execution_count": 36, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Previous chimpanzees models is in chapter 11\n", "\n", "df = pd.read_csv('./data/chimpanzees.csv', sep=';')\n", "df.head()" ] }, { "cell_type": "code", "execution_count": 37, "id": "f0597ffa", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
actorrecipientconditionblocktrialprosoc_leftchose_prosocpulled_lefttreatment
01NaN0120101
11NaN0140011
21NaN0161002
31NaN0180101
41NaN01101112
\n", "
" ], "text/plain": [ " actor recipient condition block trial prosoc_left chose_prosoc \\\n", "0 1 NaN 0 1 2 0 1 \n", "1 1 NaN 0 1 4 0 0 \n", "2 1 NaN 0 1 6 1 0 \n", "3 1 NaN 0 1 8 0 1 \n", "4 1 NaN 0 1 10 1 1 \n", "\n", " pulled_left treatment \n", "0 0 1 \n", "1 1 1 \n", "2 0 2 \n", "3 0 1 \n", "4 1 2 " ] }, "execution_count": 37, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['treatment'] = 1 + df['prosoc_left'] + 2 * df['condition']\n", "df.head()" ] }, { "cell_type": "code", "execution_count": 38, "id": "c340e064", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (1/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 1% (101/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 3% (201/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 4% (301/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 9% (700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 14% (1100/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 18% (1400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 21% (1700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 29% (2300/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 32% (2600/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 52% (4200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 69% (5500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 84% (6700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 0.000119 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.19 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 0.000119 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.19 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_45vee3sf/model_vvfe2fid.stan', line 36, column 8 to column 39)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 0.000121 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.21 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_45vee3sf/model_vvfe2fid.stan', line 36, column 8 to column 39)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 0.000161 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.61 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int N;\n", " int qty_chimpanzees;\n", " int qty_blocks;\n", " int qty_treatments;\n", " \n", " array[N] int pulled_left;\n", " array[N] int actor;\n", " array[N] int block;\n", " array[N] int treatment;\n", " }\n", " \n", " parameters {\n", " vector[qty_treatments] beta;\n", " \n", " vector[qty_chimpanzees] alpha;\n", " real bar_alpha;\n", " real sigma_alpha;\n", " \n", " vector[qty_blocks] gamma;\n", " real sigma_gamma;\n", " \n", " }\n", " \n", " model {\n", " vector[N] p;\n", " \n", " // priors\n", " beta ~ normal(0, 0.5);\n", " \n", " alpha ~ normal(bar_alpha, sigma_alpha);\n", " bar_alpha ~ normal(0, 1.5);\n", " sigma_alpha ~ exponential(1);\n", " \n", " gamma ~ normal(0, sigma_gamma);\n", " sigma_gamma ~ exponential(1);\n", " \n", " // link\n", " for (i in 1:N){\n", " p[i] = alpha[ actor[i] ] + gamma[ block[i] ] + beta[ treatment[i] ];\n", " p[i] = inv_logit(p[i]);\n", " }\n", " \n", " // linkelihood\n", " pulled_left ~ binomial(1, p);\n", " }\n", "\n", "\"\"\"\n", "\n", "dat_list = df[['pulled_left', 'actor', 'block', 'treatment']].to_dict('list')\n", "dat_list['N'] = len(df)\n", "dat_list['qty_chimpanzees'] = len(df['actor'].unique())\n", "dat_list['qty_blocks'] = len(df['block'].unique())\n", "dat_list['qty_treatments'] = len(df['treatment'].unique())\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "markdown", "id": "b29208fc", "metadata": {}, "source": [ "### R Code 13.22" ] }, { "cell_type": "code", "execution_count": 39, "id": "1d9fb8ea", "metadata": {}, "outputs": [], "source": [ "model_13_4 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list\n", ")" ] }, { "cell_type": "code", "execution_count": 40, "id": "834aa576", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
beta[0]-0.1780.296-0.5690.3520.0420.03054.01029.01.06
beta[1]0.3320.321-0.0830.8280.0770.05620.01608.01.14
beta[2]-0.5370.308-0.994-0.0690.0590.04331.01329.01.09
beta[3]0.2320.297-0.2210.7020.0460.03345.01159.01.06
alpha[0]-0.2630.405-0.8290.3200.1060.07717.01605.01.17
alpha[1]4.7631.2492.8116.4600.0350.0251182.01293.01.03
alpha[2]-0.6240.347-1.205-0.1040.0210.024400.01396.01.03
alpha[3]-0.5880.393-1.237-0.1120.0850.06124.01595.01.12
alpha[4]-0.2750.387-0.8990.2200.0920.06620.0860.01.14
alpha[5]0.6200.3530.0331.1610.0160.011507.01421.01.03
alpha[6]2.1210.4371.3832.7820.0130.0091137.01182.01.26
bar_alpha0.6780.714-0.4431.7950.0890.06365.01892.01.04
sigma_alpha2.0690.6241.0582.8310.0360.026135.01582.01.04
gamma[0]-0.1540.208-0.4710.1010.0250.01876.01477.01.04
gamma[1]0.0380.167-0.2270.2990.0040.0062537.01366.01.27
gamma[2]0.0430.173-0.2140.3190.0040.0061370.01513.01.05
gamma[3]0.0050.164-0.2720.2480.0030.0052437.01654.01.29
gamma[4]-0.0300.172-0.2860.2460.0030.0052766.01883.01.28
gamma[5]0.1030.187-0.1600.4020.0140.010395.01375.01.04
sigma_gamma0.1960.1730.0170.4050.0340.02511.04.01.29
\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "beta[0] -0.178 0.296 -0.569 0.352 0.042 0.030 54.0 \n", "beta[1] 0.332 0.321 -0.083 0.828 0.077 0.056 20.0 \n", "beta[2] -0.537 0.308 -0.994 -0.069 0.059 0.043 31.0 \n", "beta[3] 0.232 0.297 -0.221 0.702 0.046 0.033 45.0 \n", "alpha[0] -0.263 0.405 -0.829 0.320 0.106 0.077 17.0 \n", "alpha[1] 4.763 1.249 2.811 6.460 0.035 0.025 1182.0 \n", "alpha[2] -0.624 0.347 -1.205 -0.104 0.021 0.024 400.0 \n", "alpha[3] -0.588 0.393 -1.237 -0.112 0.085 0.061 24.0 \n", "alpha[4] -0.275 0.387 -0.899 0.220 0.092 0.066 20.0 \n", "alpha[5] 0.620 0.353 0.033 1.161 0.016 0.011 507.0 \n", "alpha[6] 2.121 0.437 1.383 2.782 0.013 0.009 1137.0 \n", "bar_alpha 0.678 0.714 -0.443 1.795 0.089 0.063 65.0 \n", "sigma_alpha 2.069 0.624 1.058 2.831 0.036 0.026 135.0 \n", "gamma[0] -0.154 0.208 -0.471 0.101 0.025 0.018 76.0 \n", "gamma[1] 0.038 0.167 -0.227 0.299 0.004 0.006 2537.0 \n", "gamma[2] 0.043 0.173 -0.214 0.319 0.004 0.006 1370.0 \n", "gamma[3] 0.005 0.164 -0.272 0.248 0.003 0.005 2437.0 \n", "gamma[4] -0.030 0.172 -0.286 0.246 0.003 0.005 2766.0 \n", "gamma[5] 0.103 0.187 -0.160 0.402 0.014 0.010 395.0 \n", "sigma_gamma 0.196 0.173 0.017 0.405 0.034 0.025 11.0 \n", "\n", " ess_tail r_hat \n", "beta[0] 1029.0 1.06 \n", "beta[1] 1608.0 1.14 \n", "beta[2] 1329.0 1.09 \n", "beta[3] 1159.0 1.06 \n", "alpha[0] 1605.0 1.17 \n", "alpha[1] 1293.0 1.03 \n", "alpha[2] 1396.0 1.03 \n", "alpha[3] 1595.0 1.12 \n", "alpha[4] 860.0 1.14 \n", "alpha[5] 1421.0 1.03 \n", "alpha[6] 1182.0 1.26 \n", "bar_alpha 1892.0 1.04 \n", "sigma_alpha 1582.0 1.04 \n", "gamma[0] 1477.0 1.04 \n", "gamma[1] 1366.0 1.27 \n", "gamma[2] 1513.0 1.05 \n", "gamma[3] 1654.0 1.29 \n", "gamma[4] 1883.0 1.28 \n", "gamma[5] 1375.0 1.04 \n", "sigma_gamma 4.0 1.29 " ] }, "execution_count": 40, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_4, hdi_prob=0.89)" ] }, { "cell_type": "code", "execution_count": 41, "id": "068dc7c7", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "az.plot_forest(model_13_4, hdi_prob=0.89, combined=True, figsize=(15, 8))\n", "plt.grid('--', color='white', alpha=0.2)\n", "plt.axvline(x=0, color='red', alpha=0.5, ls='--')\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 42, "id": "f923270b", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(17, 6))\n", "\n", "az.plot_dist(\n", " [model_13_4.posterior.sigma_gamma], color='blue', quantiles=[.05, .89]\n", ")\n", "\n", "az.plot_dist(\n", " [model_13_4.posterior.bar_alpha + model_13_4.posterior.sigma_alpha],\n", " color='black', quantiles=[.05, .89]\n", ")\n", "\n", "plt.legend(['Block', 'Actor'])\n", "plt.title('Posteioris')\n", "plt.ylabel('Density')\n", "plt.xlabel('Standard Deviation')\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "efa9f1c0", "metadata": {}, "source": [ "### R Code 13.23" ] }, { "cell_type": "code", "execution_count": 43, "id": "0564b736", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 2% (200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 9% (700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 15% (1200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 40% (3200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 62% (5000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 81% (6500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 7.1e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.71 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 0.000102 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.02 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_rpq3tyko/model_r7i3x47h.stan', line 28, column 8 to column 47)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 9e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.9 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 6.3e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.63 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int N;\n", " int qty_chimpanzees;\n", " int qty_blocks;\n", " int qty_treatments;\n", " \n", " array[N] int pulled_left;\n", " array[N] int actor;\n", " array[N] int block;\n", " array[N] int treatment;\n", " }\n", " \n", " parameters {\n", " vector[qty_treatments] beta;\n", " \n", " vector[qty_chimpanzees] alpha;\n", " real bar_alpha;\n", " real sigma_alpha; \n", " }\n", " \n", " model {\n", " vector[N] p;\n", " \n", " // priors\n", " beta ~ normal(0, 0.5);\n", " \n", " alpha ~ normal(bar_alpha, sigma_alpha);\n", " bar_alpha ~ normal(0, 1.5);\n", " sigma_alpha ~ exponential(1);\n", " \n", " // link\n", " for (i in 1:N){\n", " p[i] = alpha[ actor[i] ] + beta[ treatment[i] ];\n", " p[i] = inv_logit(p[i]);\n", " }\n", " \n", " // linkelihood\n", " pulled_left ~ binomial(1, p);\n", " }\n", "\n", "\"\"\"\n", "\n", "dat_list = df[['pulled_left', 'actor', 'block', 'treatment']].to_dict('list')\n", "dat_list['N'] = len(df)\n", "dat_list['qty_chimpanzees'] = len(df['actor'].unique())\n", "dat_list['qty_blocks'] = len(df['block'].unique())\n", "dat_list['qty_treatments'] = len(df['treatment'].unique())\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 44, "id": "b029a06d", "metadata": {}, "outputs": [], "source": [ "model_13_5 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list.keys()\n", ")" ] }, { "cell_type": "markdown", "id": "3b11b52a", "metadata": {}, "source": [ "### R Code 13.24" ] }, { "cell_type": "code", "execution_count": 45, "id": "46698673", "metadata": {}, "outputs": [], "source": [ "# az.compare(model_13_4, model_13_5)" ] }, { "cell_type": "markdown", "id": "389b3d40", "metadata": {}, "source": [ "### R Code 13.25" ] }, { "cell_type": "code", "execution_count": 46, "id": "e34bd922", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (1/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 1% (101/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 3% (201/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 4% (301/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 6% (500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 9% (700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 11% (900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 15% (1200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 19% (1500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 21% (1700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 24% (1900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 28% (2200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 31% (2500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 36% (2900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 55% (4400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 71% (5700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 86% (6900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 0.000122 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.22 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_59n9gjpm/model_ncxzodym.stan', line 31, column 8 to column 37)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 9.1e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.91 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 9.8e-05 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.98 seconds.\n", " Adjust your expectations accordingly!\n", " Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:\n", " Exception: normal_lpdf: Scale parameter is 0, but must be positive! (in '/tmp/httpstan_59n9gjpm/model_ncxzodym.stan', line 34, column 8 to column 47)\n", " If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,\n", " but if this warning occurs often then your model may be either severely ill-conditioned or misspecified.\n", " Gradient evaluation took 0.000137 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.37 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int N;\n", " int qty_chimpanzees;\n", " int qty_blocks;\n", " int qty_treatments;\n", " \n", " array[N] int pulled_left;\n", " array[N] int actor;\n", " array[N] int block;\n", " array[N] int treatment;\n", " }\n", " \n", " parameters {\n", " vector[qty_treatments] beta;\n", " real sigma_beta;\n", "\n", " vector[qty_chimpanzees] alpha;\n", " real bar_alpha;\n", " real sigma_alpha;\n", " \n", " vector[qty_blocks] gamma;\n", " real sigma_gamma;\n", " \n", " }\n", " \n", " model {\n", " vector[N] p;\n", " \n", " // priors\n", " beta ~ normal(0, sigma_beta);\n", " sigma_beta ~ exponential(1);\n", " \n", " alpha ~ normal(bar_alpha, sigma_alpha);\n", " bar_alpha ~ normal(0, 1.5);\n", " sigma_alpha ~ exponential(1);\n", " \n", " gamma ~ normal(0, sigma_gamma);\n", " sigma_gamma ~ exponential(1);\n", " \n", " // link\n", " for (i in 1:N){\n", " p[i] = alpha[ actor[i] ] + gamma[ block[i] ] + beta[ treatment[i] ];\n", " p[i] = inv_logit(p[i]);\n", " }\n", " \n", " // linkelihood\n", " pulled_left ~ binomial(1, p);\n", " }\n", "\n", "\"\"\"\n", "\n", "dat_list = df[['pulled_left', 'actor', 'block', 'treatment']].to_dict('list')\n", "dat_list['N'] = len(df)\n", "dat_list['qty_chimpanzees'] = len(df['actor'].unique())\n", "dat_list['qty_blocks'] = len(df['block'].unique())\n", "dat_list['qty_treatments'] = len(df['treatment'].unique())\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 47, "id": "051d5432", "metadata": {}, "outputs": [], "source": [ "model_13_6 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list.keys()\n", ")" ] }, { "cell_type": "code", "execution_count": 48, "id": "92985ff0", "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABDAAAALBCAYAAACuilGQAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAAsTAAALEwEAmpwYAAB1ZklEQVR4nOzde3ycdZ33/9enSZOmrW0K9JCyxkYqcqi0orDgYaHgcvC07N6s4lYX3K4IiMq6KoKsK3i4ZQWt3qwiWnQPLuCqt7p4V1ix8FtZFDwhhSILBAI0aUtpWtKkk0O/vz9mGkNpmzSdmetK8no+Hnlk5prr8J6ydjvv+X6/V6SUkCRJkiRJyrNJWQeQJEmSJEkajgWGJEmSJEnKPQsMSZIkSZKUexYYkiRJkiQp9ywwJEmSJElS7llgSJIkSZKk3LPAkCRJuRIRr46IWyNiQ0Q8GxG/ioi/2s1+LRHx7YjojIhtEbE6Il65D9d5V0Q8GBGFiPhdRJy3m33+IiIeiYhnIuKrEdGwy+sLShn/cITXPCciUkQs3M1rtaXXPj5k24mlbTt/eiLiyYj4fxHx1xFRt5vzPBYR/zqiPwRJksYQCwxJkpQbEXEU8GNgMvAu4M+Ae4CVEXH+kP0OBH4KLALeDZxVeml1RBw+guu8C/gK8B3gNODfgS/tco1DgX8CvgacA5wOXLzLqb4A/FtK6ef7+l730fuA44FTgL8F1gH/CNwdEbMrfG1JknKhNusAkiRJQ5wF1ABvSil1lbb9Z6nY+Evgy6Vt5wNzgT9KKT0CEBE/AR4FLgfesqcLREQt8CngX1JKHy1tXh0R84FPRMTXUkp9wB8D/5NS+t+l4w6nWKh8vPT8jcCrgJeW440PY21K6WdDnt8UESuBnwDXA2+qQgZJkjLlCAxJkpQndUAf0LPL9i08998tx1EsFx7ZuSGltA34L+CNpZJiT44HZgO7TrP4F+BA4DVDsgzN0Q1MAShNJfki8KGU0jPDv63ySyndBVxL8f0ekkUGSZKqyQJDkiTlyTdKv78YEfMjorE03eNk4PND9hsAendzfAFoAPb2gf7I0u81u2y/v/T7iNLvnwOLI+KPS6Mz3gnsHAVxKfAUxSkmo1FTWvNi8IfiyJN99f9Kv189yhySJI0ZTiGRJEm5kVJaExEnAv8XuKC0uQ84L6V045Bdfwf8cUQcmFLaBBARk4BjS68fsJfL7Hxt8y7bnxn6ekrpvyPiC8Ctpe33Ah+PiJcAfwMcl1JK+/D2hnpwlMftqq30u6lM55MkKbccgSFJknKjVA58h+JoiDcBr6M4TeLaiFg2ZNdrKf475p8j4pCIaKI4paOl9PqOcuRJKf0txekmC4GXp5TagWuAL5fKlrdGxNqI2BQRN0XE3oqTof4UOGaXn+NGETF2Rh3FsZIkjSmOwJAkSXnyaYojLt5YWkgT4LbSXUe+EBE3pJR2pJQeLRUa/wg8XNrvVxSnmXwQaN/LNXaOvJi1y347y4fnrGmRUnoaeBogIv6c4hST/1Va1POfgTdQnG7yLYp3JXnHCN7nmpTSw0M3DLNux568sPR7b+9XkqRxwREYkiQpT14G3DukvNjpbooLbM7ZuSGl9B3gYIqFwsKU0iuA6cATKaU29mznWhdH7rJ959oXD+zuoIiYTrEguah0h5TXUSwifpxSehb4EsVbslbTG0q/f1rl60qSVHUWGJIkKU86gCURUbfL9j8EtvP80REDKaW1KaVHSgttvpXf32p1T+6iOKJi2S7b3146/517OO7jwG9LxclO04Y8ns7vp3RUXEQcD5wHfC+l1Fqt60qSlBWnkEiSpDy5Bvh34D8i4ksUb2P6ZuBtwOdTSr0AETEZ+AfgDmArxdEUl1AcXXH10BNGxMPA4ymlkwFSSn0R8XfAlyLiKeDHwEnAXwHv3XmNXc6xCDgXWDJk80+AFRHxMYojRC7j9wt+ltvhEdFF8d9uTcApFKeqPAC8q0LXlCQpVywwJElSbqSUvh0RrwcuBr4GTAEeAd4DfGXorsBLgL8AGoEngeuBT++mgHjeLUpTStdGRAL+FvgQxbt5XJhS+tIeov0jcFVK6dEh57g/Is6hODLjg8BtwPv37R2P2BdLvwvAJop3RHkP8C+7K1wkSRqPYvR3/5IkSZIkSaoO18CQJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPe9CMg695jWvSZs3b846hiRJkiRJ++yBBx64JaV02q7bvQvJOLRmzRr/o0qSJEmSxqRFixbF7rY7hUSSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7LuIpSZIkSZqwUkr09/fj+pDVFxHU1tYSsdslL57HAkOSJEmSNGH19/czadIkJk2aNOIP0tp/KSV27NhBf38/kydPHtExTiGRJEmSJE1YKSXLiwxEBJMmTdqnkS8WGJIkSZKkCc3yIhv7+udugSFJkiRJknLPAkOSJEmSJOWei3iOQ+3t7RQKhaxjSJIkSVLuzZkzZ1zdgeRNb3oT//Iv/0JjY+Oo97n88sv56U9/yqxZs/jWt741uP3LX/4yd9xxB5MmTWLWrFl8/OMfZ/bs2Xu8zi9+8Qs+97nP0d/fT2NjI9ddd93z9hkYGGDdunXP2bZo0aLdns8CYxwqFApMnTo16xiSJEmSlHs7F5McT0bynva2z5vf/GbOOussPvaxjz1nn7PPPpv3vOc9ANxwww187Wtf46Mf/ehuz/Hss89y5ZVXcs0119DU1MQzzzyz2+vt2LFjxJ9fLTAkSZIkSQIOvu//0LDl4bKes2fmQp562Xv3us+6deu48MILednLXsa9997LkUceyZvf/GauvfZannnmGT71qU/xwhe+kMsvv5wnn3ySKVOmcNlll3HooYfS2dnJpZdeyoYNGzjqqKOeM5rkhz/8ITfeeCN9fX0sWrSISy65hJqammEzv+IVr3jeqAiA6dOn//599fTsdRHOVatWcdJJJ9HU1ATAAQccMOx1hzO+aiZJkiRJksagJ554gre//e1897vfpbW1lVWrVnH99dfzN3/zN1x//fVce+21vPSlL+Vb3/oWF154IR/72McAuO6661iyZAnf/va3Wbp0KR0dHQA8+uij3HrrrVx//fXceOON1NTUsGrVqv3Oec0113D66aezatUqzj///D3u9/jjj7N161be9a538Rd/8RfcfPPN+31tR2BIkiRJkgTDjpSopPnz5/OSl7wEgEMOOYRjjz2WiGDhwoWsW7eO9vZ2PvvZzwJw7LHHsmXLFrq6uvjVr37FVVddBcBrX/taZsyYAcDdd9/N2rVrecc73gEUlxqYNWvWfue88MILufDCCweLkT2VGAMDA6xdu5avfOUrbN++nXPOOYeXvexlvOhFLxr1tS0wJEmSJEnKWF1d3eDjSZMmDT6fNGkSAwMD1Nbu+8f3N73pTbz3vZUpZU4//XTe97737bHAmDt3LjNnzqShoYGGhgaOPvpoHnroof0qMJxCIkmSJElSzr385S8fnALyi1/8gsbGRqZPn87RRx/Nj370IwDuvPNOtm7dChRHafz4xz/mmWeeAWDLli27XddiX7S1tQ0+vuOOO1iwYMEe9z3hhBP4zW9+Q39/Pz09PaxZs4aWlpb9ur4jMCRpgtpSSGzs2cHshknMrN/zAkySJEnK3rvf/W4uv/xy3vKWtzBlyhQuv/xyAM4991wuvfRSzjzzTBYvXsy8efMAePGLX8wFF1zABRdcwI4dO6itreUjH/kI8+fPH/Zal1xyCb/85S/p7OzktNNO47zzzuOMM87gi1/8Io8//jgRQVNT0x7vQLLz+q961at461vfyqRJkzjjjDNYuHDhfv0ZxHi6362Kbr755uRtVCXtzV3t/dzwYD/zptfS0dXP2w6r5fgmO21JkjTxvOAFLxjV9AyVR39/P88+++xztp100km7/XbN/0qSALjynkLWEcaUp3t2sGl71ilGr64mWHH6PJob62jr7OWiVR189b6+rGNVxIFT4KCGiT1j8uJj6rOOIEmStN/GTIEREbcDa1JKF2adZajR5IqIBPx5SunbFQsmSXsxe2oNzY3FhaGaG+s4aGoN657tzziVJEmSqqWzs5PzzjvveduvvfZaGhsbR3ye73//+9xwww3P2bZ48WIuueSS/Y34PGOmwJBUWX5Du2++93AfP3h07H7g39g9QFtn7+AIjKe7B7KOVDGvnl/LGQsnZx1DkiQpVxobG7nxxhv3+zx/8id/wp/8yZ+UIdHwJnSBERGTU0rjc8y0pIo6Y+HkMf2h+K72fj58S8fgGhhnH+EaGJIkScq3sTYpuDYivhARm0s/n42ISQAR8faIuCcino2IDRHx7xFx8M4DI+LEiEgR8fqIuDsieoFT93axiDgkIr4fER0RsS0ifhURbxzmmMci4uMR8a8R0VU69oO72fWAUsZtEfFoRLx9l/N8JiJ+FxE9pXP+Q0RMGfkflSTt2fFNtXziVfW85SXBJ15Vb3khSZKk3BtrBcYyipmPB94NnAtcVHqtDvh7YDHwRuAg4Ibnn4IrgcuAw4CfD3O96cAq4I9L5/0O8N2IOGyY4z4ArAWOLmX6dET82S77fAz4fum8NwHXR0TzkNe3AX8FHA5cAJwF7PkeNZK0j2bWBwsba7yFqiRJksaEMXMb1dJimfOBl6ZS6Ii4DDgvpfQHu9n/MIolwgtTSk9GxInAauDMlNJ39iPHz4CbU0qfHJJrcBHPiHgM+J+U0h8POeZrwGEppdeUnifgMymlS0rPa4GtwLkppX/dw3XPAz6YUhr2xrneRlWSJEmSRsbbqGZrX26jOtZGYPwsPbdxuQs4OCJmRMTRpekej0fEs8AvSvs073KOXzBCETGtNHXjgdKUlS7glbs5567u2s3zI3bZ9tudD1JK/cBGYM6Qa58ZET8tTUHpAj4/gutKmsDqejczY+uD1PVuzjqKJEmSMvKGN7yBzZv3/u/B4fb5+Mc/zsknn8yf//mfP2f7l770Jd7ylrdw1llnccEFF7Bx48Zh89x///0cc8wx/PjHPx7ZG9iLsVZg7EkAtwDdwDuAY4DTSq/V7bLvtn0471XAnwN/B5wALAHu3s05R2PXxUMTpf8eEXEccCPF9/Qm4OUUp72M3RUDJVXU3PV38Ie/fD+HP/YN/vCX72fu+juyjiRJkqQx6k1vehPXXHPN87b/5V/+Jd/61re48cYbee1rX8t111231/MMDAzwhS98geOOO64sucbaOJk/jIgYMgrjOGAdsJDimheXppRaAXaz5sRovAb4551TTkqLaB4CPDTMcbv+1zmO4nSWkXo18FRK6RM7N0TEi/bheE1AL7/3sqwjjGlTtm+goTB8g5xXO2rqeeTklRRmtFC/tZXDb1vOkb9bkXWssuipn832KXOG33GC+/XiT2YdQZKkMW/lr7bQ2lneG1W2NE5m+dEz97rPunXruPDCC3nZy17Gvffey5FHHsmb3/xmrr32Wp555hk+9alP8cIXvpDLL7+cJ598kilTpnDZZZdx6KGH0tnZyaWXXsqGDRs46qijGDpp4Yc//CE33ngjfX19LFq0iEsuuYSampphM7/iFa9g3bp1z9s+ffr0wcc9PT1E7H0ttRtvvJGTTz6ZBx54YNhrjsRYG4ExH1gRES+NiDOBD1GcWtEGFIALI+LFEfEG4BN7Oc9IPQT8aWl6ysuAfwVGcieQ4yLikoh4SUS8C/jLUs59ue7BEbGs9H7OB962z+klTRh9DXMozGgBoDCjhb4GP/BLkiSNJU888QRvf/vb+e53v0trayurVq3i+uuv52/+5m+4/vrrufbaa3npS1/Kt771LS688EI+9rGPAXDdddexZMkSvv3tb7N06VI6OjoAePTRR7n11lu5/vrrufHGG6mpqWHVqlX7nfOaa67h9NNPZ9WqVZx//vl73G/Dhg2sXr36edNQ9sdYG4HxTaCG4t1DErAS+HxKaSAizgY+DbyH4voSHwB+tJ/X+0DpGv8FbAZWMLIC43PAURTvGrIN+FhK6dsjvWhK6T8i4rOl6zUAt1K8a8mX9iG7Jhi/fd0/LY/dSEvbTVnHGLXJPRuo39o6OAJjcs+GrCOVTcfck2hdcFbWMSRJ0gQw3EiJSpo/fz4veclLADjkkEM49thjiQgWLlzIunXraG9v57Of/SwAxx57LFu2bKGrq4tf/epXXHXVVQC89rWvZcaMGQDcfffdrF27lne84x0AFAoFZs2atd85L7zwQi688MLBYmRPJcZVV13F+973PiZNKt+4iTFTYKSUThzy9MLdvH4TxduRDhVDXr996PMRXvNx4HW7bL5qL7l26kop7XHERErpeTlSSgt2eX4JcMkuu315L3El7YfWBWeN6Q/Jc9ffwaGrz6d32nzqtq1j7cLzWT/3hKxjSZIkaYTq6n6/1OKkSZMGn0+aNImBgYFR3SnlTW96E+9973vLlnGo008/nfe97317LDAeeOABLrmk+JG2s7OTn/70p9TU1LB06dJRX3OsTSGRJO3G+rkn8PNXfIG1C87h56/4guWFJEnSOPPyl798cArIL37xCxobG5k+fTpHH300P/pRcfLBnXfeydatW4HiKI0f//jHPPPMMwBs2bJlt+ta7Iu2trbBx3fccQcLFizY474333wzP/zhD/nhD3/I6173Oi655JL9Ki9gDI3AqISIuB/Y0+KY704pfbOaeSRpf/TWzaK3bv+HBUqSJCl/3v3ud3P55Zfzlre8hSlTpnD55ZcDcO6553LppZdy5plnsnjxYubNmwfAi1/8Yi644AIuuOACduzYQW1tLR/5yEeYP3/+sNe65JJL+OUvf0lnZyennXYa5513HmeccQZf/OIXefzxx4kImpqa+OhHP1rR97yrGLpC6URTurPHnm5Nuj6l9Gw185TLzTffnKZOnZp1DEmSJEnKvRe84AWjmp6h8ujv7+fZZ5/70fukk07a7fIPE/q/UmmNC0mSJEmSlHMTusCQJEmSJGki6uzs5Lzzznve9muvvZbGxsYRn+f73/8+N9xww3O2LV68eHABz3Ka0FNIxiunkEiSJEnSyLzgBS+gpqaGiH26aaXKIKXEwMDAiKeQeBcSSZIkSdKENTAwABQ/TKt6dv557/zzHwmnkEiSJEmSJqyenh4AampqMk4y8QwMDAz++Y+EBYYkSZIkacJKKdHd3Z11DI2AU0gkSZIkSVLuWWBIkiRJkqTcs8CQJEmSJEm55xoY41B9fb1zuCRJkiRJ44oFxjjU1NSUdQRJkiRJksrKKSSSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTcs8CQJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYYkiRJkiQp9ywwJEmSJElS7llgSJIkSZKk3LPAkKQhNvcMsHZjgc09A1lHkSRJkjREbdYBJCkvVrd2cd09m5n3glo6nu3n3GNmsbRletaxJEmSJLGfBUZEnAisBmanlJ4uRyBJ2l8fubVjt9vXd/WzsXvPIyvqaoIVp8+jubGOts5eLlrVwdV3bhrRNWdPrWHu9H3/K/Uzp8zb52MkSZKkiWh/p5D8N9AEjOxf+JKUY7On1tDcWAdAc2MdB02tyTiRJEmSpJ32awRGSqkX2P1XnZKUkT2NavjmvZ3ccN+WPR63sXuAts7ewREYT+9ltMauXnfIdJYtbtzXqJIkSZJGKFJKw+8U8UfAPwCLgAHgd8BfAQexyxSSiPgr4OOl124DVgH/mFKK0usfB84EPgtcDswGvgW8G/hr4BJgKvBPwAdTSjtKx70deD9wGNAD3AFclFJ6akRvNOINwOeAFwF3A18CbgBaUkqPRcSBwDXAa4EDgUeBq1JKXx9yjtuBtUA38M7Sn8UngWtL514GbAU+mlL6l9IxC4BW4G3A+cCxwIPA2cAO4DpgMfBr4B0ppdbScYeUzvmHwAtKf+YfSyndPNx7XbNmzfD/USU9j2tgSJIkSdlbtGhR7G77sCMwIqIW+D6wkuIH9MnA0RQ/vO+67/HA1yiWEP8XOAH49G5OuwD4E+CNwMHAdyhORWkHTqFYUnwLuLP0GkAd8PcUP/wfBFxJsYD4oxG8h2bgu8A/Al8BXkaxHBhqCvCr0nm3Aq8DvhIRbSml24bst4zfFwtvBlYApwE/Al5JsZj4WkT8OKXUPuS4y4G/oViMfLmUfQPw0dLvfwK+CLyptP90iuXPZRQLm7cC342Io1JKDw73niXtu6Ut01kyr4GOrn7mTa9lVoNTSCRJkqS8GHYERkQcQHGNixNTSnfs8tqJDBmBERE3ALNSSqcN2ec64F27jMC4GJiXUtpS2vZtimXHwaVpKTtHO6xJKV24h1yHURwN8cKU0pPDvIf/DZyRUjp8yLZLgU9RGoGxh+NuBLpSSn89JFN9Sun40vOgWD7clVJ6c2nbZGAb8BcppW8PGYFxXkrpK6V93gj8B/C/UkrfLW07B7gmpbTHr3sj4mfAzSmlT+7t/ToCQ5IkSZI0Vu1pBMawi3imlJ4BvgHcEhE/jIgPlEY07M5hFKdnDPXz3ezXtrO8KFkPPLSzvBiybc7OJxFxdER8PyIej4hngV+UXtpTll1z3bO3XBFRExEfjYjfRsSmiOgC/mw35//tzgep2P5sAO4bsq0P2Dw0+67Hld4bQ48rbZsWEVNLeaZFxD9ExAMRsbmU55W7ySOpAmq3b6Jh0xpqt7tGsSRJkpQHI7oLSUrpnRSnTPx/FKdN/C4iTt2P6/bteok9bJsExQ/zwC0U1554B3AMxWkbUJxaUg4fBP6W4tocJwNLgO/t5vz7lH0Px6W9bNt53FXAnwN/R3F0yhKK5VC53q+kPZjZdgsLb307Tb9ZwcJb387MtluyjiRJkiRNeCO+C0lK6V7gXuDKiFhFca2H63bZ7UGK5cJQx+5XwqLDKK57cemQRS7/bB+Of5Dimht7y/Ua4D+GLL4ZwKFA52gCl8FrgH9OKX2nlGcKcAjwUEZ5pDGj5fbdzjwbNHlbO3U9e76B0o6aeh45eSWFGS3Ub23lkNuW88K7rxj2ur0N8+ib1rRPWVtPvGaf9pckSZImqmFHYERES0R8JiJeFREvioilwFHAA7vZ/YvAKRHxoYh4SUQsB/60DDnbgAJwYUS8uHRHkU/sw/HXAodExFUR8dJS+fHu0ms7Rz48BJwcEa8pra9xDdBShuyj9RDwp6WpMy8D/pXiQqOSKqyvYQ6FGcX/+RdmtNDXsOuMMEmSJEnVNpIRGN0URyL8O8VREOuBb1K8W8erh+6YUrorIt5F8Y4bVwA/Lu2310Unh5NS2hgRZ1O8o8l7KK4n8QGKd/4YyfGPR8T/onj3kAsprodxOXA9sL202ycpFharKN714xsU3+cR+5N9P3yA4p1f/ovimhorsMCQRmS4UQ1z7l/JnLXX7/H1yT0bqN/aOjgCY3LPhhFdt3PB69lw5PJ9yipJkiRpZIa9C8l+XyDi88DrUkovq+iF9lFEvJ9iydKYKv2HUGXehUTaPzPbbqHpNyvonXYwddueon3JRWxp3p9lfyRJkiSN1J7uQjLiNTBGKiI+BPwn0AW8DjgPuLTc19lXEfEeiiMvNgLHUVwc8xvjrbyQtP+2NJ/KtjmvZPK2dvqmNdE/5cCsI0mSJEkTXtkLDIq3+vwgMBNoBS4BvlCB6wyKiGuBt+/h5X9NKZ0HLKRYpBwIPElxXYzhV+WTNCH1TznQ4kKSJEnKkYpPIamGiJgDzNjDy1tTSiObwD5OOIVEkiRJkjRWVW0KSRZKBcWEKikkSZIkSZpIhr2NqiRJkiRJUtYsMCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3arMOoPJrb2+nUChkHUOSJEmSpH22aNGi3W63wBiHCoUCU6dOzTqGJEmSJEll4xQSSZIkSZKUexYYkiRJkiQp9ywwJEmSJElS7llgSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs8CQ5IkSZIk5Z4FhiRJkiRJyj0LDEmSJEmSlHsWGJIkSZIkKfcsMCSVzZZC4uHOAbYUUtZRJEmSJI0ztVkHkDQ+3NXezw0P9jNvei0dXQXedlgtxzf5V4wkSZKk8vDThVRhV95TyOS6T/fsYNP26l2vriZYcfo8mhvraOvs5aJVHXz1vr7qBdiNA6fAQQ3ZDTS7+Jj6zK4tSZIkjTcTcgpJRKSIOHMf9j+xdMxB+3nd20vnSRFx3CiunyLi5v3JIFXK7Kk1NDfWAdDcWMdBU2syTiRJkiRpPHEERvV9HbgU2LRzQ0TMAr4IvLm06QfAe1NKnaXn/w00AV8AplUtqcoiq2/hv/dwHz94tL9q19vYPUBbZ+/gCIynuweqdu09efX8Ws5YODnrGJIkSZLKwAKj+rpTSh27bPs3oBk4rfT8a8C/AG8CSCn1Ah0R0YMFhkbojIWTq/rh/a72fj58S0dpDYx+zj7CNTAkSZIklc+4nEISEadFxH9FxOaIeCYibomIw/ew74LS1Iy/iIifRsT2iHgwIk7Zze6LI+LnEdEdEb+IiKOHnOfAiLghIp6MiJ6IuD8i3jmCrIdTLC7OTSndlVK6C3g38MaIeOko/wikqju+qZZPvKqet7wk+MSr6i0vJEmSJJXVuCwwKI5SWAEcC5wIbAH+IyLq9nLMP1CcxrEE+E/g+xFx8C77/G/gI8DRFKeAfDMiovTaFOBXwBuBIylO9/hKRJw8TNbjgS6K00R2uhPYBrxqmGOlXJlZHyxsrGFmfQy/syRJkiTtg3H5FWlK6TtDn5dGQmylWGj8dA+HfTml9K3S/u8HTgXOBy4bss/fpZRWl/a5onSug4EnU0pPAZ8dsu91EXES8Dbgtr3EnQdsTCmlIflTRGwovSZJkiRJ0oQ3LkdgRMQhEfFvEfFIRGwF1lN8r817OeyunQ9SSjuAnwNH7LLPb4c8Xlf6Pad0zZqI+GhE/DYiNkVEF/Bnw1xTGlPqejczY+uD1PVuzjqKJEmSpAlmXI7AAG4GnqS4lsRTQD/wALC3KSQj0Tfk8c4REztLoA8Cfwu8H7iP4rSQT1MqOPaiA5gdEbFzFEZpWsqc0mtSLsxdfweHPrqS3mnzqdu2jodevJz1c0/IOpYkSZKkCWLcFRgRcSBwGHDBkOkeRzP8ez0O+Elp/6A43eTb+3Dp1wD/kVL6lyHnOBToHOa4u4DpFNfC2LkOxvEU1/H47z0dpMp7+b2XDb9ThqZs30BDYWPVrrejpp5HTl5JYUYL9VtbOfy25Rz5uxUVu15P/Wy2Txmu/8vOrxd/MusIkiRJ0oQyHqeQbAaeBt4VEQsj4gTgWoqjMPbm/Ig4s3TnjxXAi4Av78N1HwJOjojXRMRhwDVAy3AHpZTWAj+iuODn8RFxPPAV4OaU0u/24fpSRfU1zKEwo/h/0oUZLfQ15LdckCRJkjT+jLsRGCmlHRHxVop3FFkDPExxasd39npg8e4iH6B4h5HHgT9NKT25D5f+JMXCYhXQA3wD+CbPX0djd/4C+D/ALaXnPwAu3IdrqwLy/g17y2M30tJ2U9WuN7lnA/VbWwdHYEzu2VDR63XMPYnWBWdV9BqSJEmSxo5xV2AApJR+AizaZfP0Ia/v7h6Pv0sp7fa2pSml24HYZdtjQ7ellDZTXLRzNHk3A28fzbGauFoXnFXVD/hz19/BoavPH1wDY+3C810DQ5IkSVLVjMsCI+fOjYhzgKUppXtGckBEvJbiyI56fj9KQ6qq9XNPYPOso5iyfT3bp8ylt25W1pEkSZIkTSAWGNW1DGgoPX5iH477BbCk9HhbOQNJ+6K3bpbFhSRJkqRMTPgCY9epIBW+1lOjPK6H4loekiRJkiRNSOPxLiSSJEmSJGmcscCQJEmSJEm5Z4EhSZIkSZJyb8KvgTEe1dfX093dnXUMSZIkSZLKxgJjHGpqaso6giRJkiRJZeUUEkmSJEmSlHsWGJIkSZIkKfcsMCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRp3NncM8DajQU29wxkHUVSmXgbVUmSJEnjyurWLq67ZzPzXlBLx7P9nHvMLJa2TM86lqT9ZIFRRRFxO3BC6enxKaWfjeCYBUBr6en9KaVFlUknSZLy6CO3dmQdQVWyvqufjd2OFiiHuppgxenzaG6so62zl4tWdXD1nZuyjlVxs6fWMHe6H/G0e585ZV7WEfabU0iq7+tAE/BLgIj4aETcGRHbIiLtZv8nSvtfXcWMkiRJ0pg1e2oNzY11ADQ31nHQ1JqME0kqB+u56utOKQ39KqUe+C5wO3DprjunlAaAjojoqk48SZKUJ+PhGzONzDfv7eSG+7ZkHWNc2Ng9QFtn7+AIjKcnyMiW1x0ynWWLG7OOIVWMBUbGUkofA4iIM7POIkmSpOwsW9zoh88yWd3axcW3rh9cA+O9xx3gGhjSOGCBIUmSJGlcWdoynSXzGujo6mfe9FpmNTiFRBoPLDAkSZIkjTuzGmosLqRxxkU8JUmSJFVV7fZNNGxaQ+328X9nEEnl4wgMSZIkSVUzs+0Wmn6zgt5pB1O37Snal1zEluZTs44laQywwJAkSaqSltsvzDqCKmTytnbqejqG31HsqKnnkZNXUpjRQv3WVg65bTkvvPuKrGNNCL0N8+ib1pR1DFVY64nXZB2hYpxCkrGIaI6IJcCC0vMlpR+XSZYkSdK409cwh8KMFgAKM1roa5iTcSJJY4UjMLJ3BXD2kOe/Lv1eCtxe9TSSJKlixvO3YhPdnPtXMmft9VnHGBMm92ygfmvr4AiMyT0bso40YXQueD0bjlyedQxp1CwwMpZSOgc4J+MYkiRJ2g8bjlzuB8MRmtl2Cy23XzC4BsZTr7jYNTAkjYgFRvWdGxHnAEtTSvcMt3NENAMPAHXAQxXOJkmSJFXUluZT2TbnlUze1k7ftCb6pxyYdSRJY4QFRnUtAxpKj58Y4THrgCWlx4VyB5IkSZKqrX/KgRYXkvaZBUYVpZSeGsUx/cDDFYgjSZIkSdKY4V1IJEmSJElS7llgSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs9FPMeh9vZ2CgVvWCJJkiRJGnsWLVq02+0WGONQoVBg6tSpWceQJEmSJKlsnEIiSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYY0jixpZB4uHOALYWUdRRJkiRJKjvvQiKNA3e193PDg/3Mm15LR1eBtx1Wy/FN/s9bkiRJ0vjhJxyNWVfeU8g6wl493bODTdurc626mmDF6fNobqyjrbOXi1Z18NX7+vb5PAdOgYMa8jUw6+Jj6rOOIEmSJCkH8vVJZZyLiNsjIpV+jhvhMQuGHLOm0hk1Ns2eWkNzYx0AzY11HDS1JuNEkiRJklRejsCovq8DlwKbImIB8HfAUqAJaAduAq5IKfWU9n+i9NoHgdOqnjbH8v7N/Pce7uMHj/ZX5Vobuwdo6+wdHIHxdPfAqM7z6vm1nLFwcpnTSZIkSdL+s8Covu6UUgdARBwG1ADnA/8DHA5cBxwInAuQUhoAOiKiK5u4Gq0zFk6uWhlwV3s/H76lo7QGRj9nH+EaGJIkSZLGFz/hZCil9CPgR0M2PRoRnwI+QanAkEbi+KZajjigho09O5jdUM/M+sg6kiRJkiSVlQVG/swANmcdQmPPzPpgZr1rX0iSJEkan1zEM0ci4kUU17r4UtZZJEmSJEnKEwuMnIiIuRSnk/wn8PmM4yhn6no3M2Prg9T1OjhHkiRJ0sTkFJIciIh5wE+ANcA7Ukop40jKkbnr7+DQR1fSO20+ddvW8dCLl7N+7glZx5IkSZKkqrLAyFhENAGrgfuBt6WUqnPfzZx6+b2XZR1hRKZs30BDYWNVrrWjpp5HTl5JYUYL9VtbOfy25Rz5uxUVu15P/Wy2T5lTsfOXy68XfzLrCJIkSZKqyCkkGYqI+cAdQAdwEXBQRMwr/bgaowDoa5hDYUYLAIUZLfQ15L9ckCRJkqRycwRGtk4BXlL6advltRbgsWoHytpY+Va95bEbaWm7qSrXmtyzgfqtrYMjMCb3bKjo9TrmnkTrgrMqeg1JkiRJ2lcWGBlKKX0D+EbGMTQKrQvOqtqH/Lnr7+DQ1ecProGxduH5roEhSZIkacKxwKi+cyPiHGBpSume4XaOiGbgAaAOeKjC2ZRD6+eewOZZRzFl+3q2T5lLb92srCNJkiRJUtVZYFTXMqCh9PiJER6zDlhSelwodyCNDb11sywuJEmSJE1oFhhVlFJ6ahTH9AMPVyCOJEmSJEljhnchkSRJkiRJuWeBIUmSJEmScs8CQ5IkSZIk5Z5rYIxD9fX1dHd3Zx1DkiRJkqSyscAYh5qamrKOIEmSJElSWTmFRJIkSZIk5Z4FhiRJkiRJyj0LDEmSJEmSlHsWGJIkSZIkKfcsMKQxbnPPAGs3FtjcM5B1FEmSJEmqGO9CIo1hq1u7uO6ezcx7QS0dz/Zz7jGzWNoyPetYkiRJklR2FhgaMz5ya0fWEYa1vqufjd3VGwlRVxOsOH0ezY11tHX2ctGqDq6+c1PVrr83s6fWMHf6/v0V85lT5pUpjSRJkqSxzikkVRQRt0dEKv0cN8JjFgw5Zk2lM2psmT21hubGOgCaG+s4aGpNxokkSZIkqTIcgVF9XwcuBTZFxCTge8ASYA6wGbgNuDil9FRp/yeAJuCDwGnVDpsnY+Hb+G/e28kN922p2vU2dg/Q1tk7OALj6SqO/hjO6w6ZzrLFjVnHkCRJkjROWGBUX3dKqQOgVGD8BPg00A4cDFwF/F/gWICU0gDQERFd2cTVvli2uLGqH9pXt3Zx8a3rB9fAeO9xB7gGhiRJkqRxyQIjQymlHcCKIZsej4jPAN+PiCkppe3ZJNNYsbRlOkvmNdDR1c+86bXManAKiSRJkqTxyQIjRyLiAGAZ8HPLC43UrIYaiwtJkiRJ456LeOZARFwZEduATUAz8MaMI0mSJEmSlCsWGPnwWeDlwCnAAPCvERHZRlKe1G7fRMOmNdRuz8ctUiVJkiSp2pxCkgMppaeBp4GHImItxTuPvAb4r0yDKRdmtt1C029W0DvtYOq2PUX7kovY0nxq1rEkSZIkqaosMPJn56iY+kxTVEnL7RdmHWHUJm9rp66no+LX2VFTzyMnr6Qwo4X6ra0ccttyXnj3FRW9Zm/DPPqmNVX0GqPVeuI1WUeQJEmSlAELjAxFxPHA0cBPgU7gEOATwGOlbRJ9DXMozGgBoDCjhb6GOdR3PZFxKkmSJEmqLguMbPUAZwJXANOAduBHwFsnyl1IxvK36XPuX8mctddX/DqTezZQv7V1cATG5J4NFb9m54LXs+HI5RW/jiRJkiSNlAVGhlJKvwGWZp1Do7PhyOVV+ZA/s+0WWm6/YHANjKdecbFrYEiSJEmacCwwqu/ciDgHWJpSume4nSOiGXgAqAMeqnA25dCW5lPZNueVTN7WTt+0JvqnHJh1JEmSJEmqOguM6loGNJQej3QRg3XAktLjQrkDaWzon3KgxYUkSZKkCc0Co4pSSk+N4ph+4OEKxJEkSZIkacyYNPwukiRJkiRJ2bLAkCRJkiRJuWeBIUmSJEmScs81MMah9vZ2CgXX+5QkSZIkjT2LFi3a7XYLjHGoUCgwderUrGNIkiRJklQ2TiGRJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYYkspiSyHxcOcAWwop6yiSJEmSxiFvoyppv93V3s8ND/Yzb3otHV0F3nZYLcc3+deLJEmSpPLxE4a0n668p5B1hN16umcHm7ZX51p1NcGK0+fR3FhHW2cvF63q4Kv39VXn4sM4cAoc1FC+wWYXH1NftnNJkiRJGjmnkFRRRNweEan0c9wIj1kw5Jg1lc4ojcbsqTU0N9YB0NxYx0FTazJOJEmSJGm8cQRG9X0duBTYNHRjREwBfg4cBRyTUvpF6aUngCbgg8BpVcypEcrrN/Lfe7iPHzzaX5VrbeweoK2zd3AExtPdA1W57ki8en4tZyycnHUMSZIkSfvJAqP6ulNKHbvZfhXwJMUCY1BKaQDoiIiuaoTT+HHGwslV++B+V3s/H76lo7QGRj9nH+EaGJIkSZLKy08YORARfwIsBc4EXp9xHGmfHd9UyxEH1LCxZwezG+qZWR9ZR5IkSZI0zlhgZCwi/gD4MnA60JNxHGnUZtYHM+td+0KSJElSZbiIZ4Yiogb4JnB1SunerPNIw6nr3cyMrQ9S17s56yiSJEmSJhhHYGTrUqAX+FzWQaThzF1/B4c+upLeafOp27aOh168nPVzT8g6liRJkqQJwgIjWycDrwX6Ip6zZsDPIuKmlNKybGJpVy+/97KsI+zWlO0baChsrMq1dtTU88jJKynMaKF+ayuH37acI3+3oiLX6qmfzfYpcypy7nL69eJPZh1BkiRJmjAsMLL1TmDakOfzgVuAZcCdmSSS9qCvYQ6FGS0AFGa00Ncwh/quJzJOJUmSJGmisMDIUEqpdejzIbdKfSSl9GQGkbQHef2mveWxG2lpu6kq15rcs4H6ra2DIzAm92yo2LU65p5E64KzKnZ+SZIkSWOPBYY0hrUuOKtqH/Tnrr+DQ1efP7gGxtqF57sGhiRJkqSqscDIkZTSY0AMt5+UhfVzT2DzrKOYsn0926fMpbduVtaRJEmSJE0g3ka1+s6NiK6IOGYkO0dEc2lqyaUVziUNq7duFltnHGZ5IUmSJKnqHIFRXcuAhtLjka5+uA5YUnpcKHcgSZIkSZLGAguMKkopPTWKY/qBhysQR5IkSZKkMcMpJJIkSZIkKfcsMCRJkiRJUu5ZYEiSJEmSpNxzDYxxqL6+nu7u7qxjSJIkSZJUNhYY41BTU1PWESRJkiRJKiunkEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkTRCbewZYu7HA5p6BrKNIkiRJ+8y7kEjSBLC6tYvr7tnMvBfU0vFsP+ceM4ulLdOzjiVJkiSNmAWGpHHrI7d2ZB1ht9Z39bOxu7qjIOpqghWnz6O5sY62zl4uWtXB1Xdu2ufzzJ5aw9zp+f1/HZ85ZV7WESRJklQhTiGpooi4PSJS6ee4ER6zYMgxayqdUdL4NHtqDc2NdQA0N9Zx0NSajBNJkiRJ+ya/X6ONX18HLgU2AUTEY8CLdtnnypTSR0qPnwCagA8Cp1UpozQu5PXb+G/e28kN922p6jU3dg/Q1tk7OALj6VGOAHndIdNZtrixvOEkSZKkEbDAqL7ulNKu49qvAL485HnXzgcppQGgIyK6kDQuLFvcWPUSYHVrFxffun5wDYz3HneAa2BIkiRpTLHAyIdnd1NqSFLZLG2ZzpJ5DXR09TNvei2zGpxCIkmSpLHFNTDy4YMRsSkifhMRH42IuqwDSRp/ZjXUcPjsessLSZIkjUmOwMjeF4FfU1wT41jgM0AL8NdZhpIkSZIkKU8sMDKWUvrckKe/jYitwE0RcXFKad/vcShpQqrdvonJ29rpm9ZE/5QDs44jSZIklZ0FRv78vPR7IaU7lUjS3sxsu4Wm36ygd9rB1G17ivYlF7Gl+dSsY0mSJEllZYGRP0tKv9uzDCGNRS23X5h1hOeYvK2dup7Kr8+7o6aeR05eSWFGC/VbWznktuW88O4rKn7d3oZ59E1rqvh1yqH1xGuyjiBJkqT9ZIGRoYg4HjgOWA1sAY4BPg/8IKXUlmU2SWNHX8McCjNaACjMaKGvYQ71XU9knEqSJEkqLwuMbBWAtwJ/D9QDjwNfBf4hy1DSWJW3b9nn3L+SOWuvr/h1JvdsoH5r6+AIjMk9Gyp+TYDOBa9nw5HLq3ItSZIkyQIjQymlX1EcgSFpHNpw5PKqfMCf2XYLLbdfMLgGxlOvuNg1MCRJkjTuWGBU37kRcQ6wNKV0z3A7R0Qz8ABQBzxU4WySxqAtzaeybc4rvQuJJEmSxjULjOpaBjSUHo90gvo6fr+wZ6HcgSSND/1TDrS4kCRJ0rhmgVFFKaWnRnFMP/BwBeJIkiRJkjRmTMo6gCRJkiRJ0nAsMCRJkiRJUu5ZYEiSJEmSpNxzDYxxqL29nULB9T4lSZIkSWPPokWLdrvdAmMcKhQKTJ06NesYkiRJkiSVjVNIJEmSJElS7llgSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs8CQ5IkSZIk5Z4FhiRJqqgthcTDnQNsKaSso0iSpDHM26hKkqSKuau9nxse7Gfe9Fo6ugq87bBajm/ynx+SJGnf+S+IKoqI24ETSk+PTyn9bATHLABaS0/vTyktqkw6SdJwrrynkHWEsnq6Zwebtlf2GnU1wYrT59HcWEdbZy8Xrergq/f1VfaiQxw4BQ5qGHsDTi8+pj7rCJIk5c7Y+//oY9/XgSbglzs3RMSpEXFXRHRHRGdE/GTI/k+U9r+6yjklSdpvs6fW0NxYB0BzYx0HTa3JOJEkSRqrHIFRfd0ppY6dTyLiDIqlxkeBcyiWSkfvfD2lNAB0RERXdWNKknY13r4V/97Dffzg0f6KXmNj9wBtnb2DIzCe7h6o6PV29er5tZyxcHJVrylJkirDAiNDEVEDfBH4cErpq0NeWptRJEnSBHLGwskV/3B/V3s/H76lo7QGRj9nH+EaGJIkaXT8F0S2XgG8EOiNiF8B84HfAhenlH6daTJJksrg+KZajjigho09O5jdUM/M+sg6kiRJGqNcAyNbLy79vgL4NPAG4Eng9ohoyiyVJEllNLM+WNhYY3khSZL2iwVGtnb++X8qpfTtlNIvgXOBLcBfZhdLkqTyqevdzIytD1LXuznrKJIkaQxzCkm22ku/H9i5IaXUHxH/AzRnE0mSpPKZu/4ODn10Jb3T5lO3bR0PvXg56+eeMPyBkiRJu7DAyNYvgQLwUuCnABExCTgEuCXDXJI04b383suyjrBfpmzfQENhY9Yx2FFTzyMnr6Qwo4X6ra0cfttyjvzdiqxj0VM/m+1T5mQdY9R+vfiTWUeQJKnqLDAylFLaGhHXApdHxJPAY8CFwCzgX7LMJklSOfQ1zKEwowWAwowW+hrmUN/1RMapJEnSWGSBkb0PAb3APwFTgV8BS1NK7Xs9SpJUUWP9G+6Wx26kpe2mrGMwuWcD9VtbB0dgTO7ZkHUkADrmnkTrgrOyjiFJkvaBBUbGUkp9wIdLP5IklUXrgrNy8QF97vo7OHT1+YNrYKxdeL5rYEiSpFGxwKi+cyPiHIqjLO4ZbueIaKa4yGcd8FCFs0mSVFbr557A5llHMWX7erZPmUtv3aysI0mSpDHKAqO6lgENpccjnQC8DlhSelwodyBJkiqtt26WxYUkSdpvFhhVlFJ6ahTH9AMPVyCOJEmSJEljxqSsA0iSJEmSJA3HAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTcs8CQJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlXm3WAVR+9fX1dHd3Zx1DkiRJkqSyscAYh5qamrKOIEmSJElSWTmFRJIkSZIk5Z4FhiRJkiRJyj0LDEmSJEmSlHsWGJIkSZIkKfcsMCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJVbe5Z4C1Gwts7hnIOookSZKkMaI26wCSJpbVrV1cd89m5r2glo5n+zn3mFksbZmedSxJkiRJOWeBIY0zH7m1Y7+OX9/Vz8buyo2MqKsJVpw+j+bGOto6e7loVQdX37mp7NeZPbWGudPz81fcZ06Zl3UESZIkaUybcFNIIuL2iLgmg+suiIhU+nlwH4/9xpBjz6xURqkaZk+tobmxDoDmxjoOmlqTcSJJkiRJY0F+vp4cIyLiRGA1MDul9PQoTnEa8MtdznkC8DngSGAd8A8ppWuH7PJ+4CNA+yiupwlmf7/p/+a9ndxw35YypXm+jd0DtHX2Do7AeLpCoz1ed8h0li1urMi5JUmSJFWfBUb1bRpafEREC/D/gOuBtwOvAb4UERtTSt8BSCltAbZERBZ5NcEsW9xY0Q/+q1u7uPjW9YNrYLz3uANcA0OSJEnSsCbcFJKS2oj4QkRsLv18NiImAUREXURcGRFPRkR3RNwTEaeWXltAcfQFwMbSlI5vlF47LSL+q3S+ZyLilog4fARZzgPWpZTem1Jam1L6KvBPwAfL/aalPFjaMp0vvWk+577yAL70pvmWF5IkSZJGZKIWGMsovvfjgXcD5wIXlV77OnAC8BfAIoplwn9ExGLgCeB/lfY7EmiiOL0DYBqwAjgWOBHYUjqubpgsxwO37rLtFuCVETF5n9+ZNAbMaqjh8Nn1zGpw/QtJkiRJIzNRp5C0A+9LKSXgwYg4FPhARHwfeBuwIKXUVtr3moh4HfDulNIFEfFMafuGoVNBdk732Cki3glspVho/HQvWeYBP95l23qK/20OwnUvJEmSJEmasCMwflYqL3a6CziY4voTATwQEV07f4A3AIfs7YQRcUhE/FtEPBIRWymWEJOA5sq8BWl8q92+iYZNa6jdXv5brEqSJEkaeybqCIy9ScAxQN8u23uGOe5m4EmKU1KeAvqBB4DhppB0AHN32Ta3dPxo7nIijXkz226h6Tcr6J12MHXbnqJ9yUVsaT4161iSJEmSMjRRC4w/jIgYMgrjOIq3L72L4giMeSml1Xs4trf0e3DyfkQcCBwGXLDzuIg4mpH9+d4F/Oku2/4Y+EVKadcSRRrUcvuFWUcAYPK2dup6Osp6zh019Txy8koKM1qo39rKIbct54V3X1GWc/c2zKNvWlNZzlVurSdek3UESZIkKbcm6hSS+cCKiHhpRJwJfAj4fErpIeCbwDci4syIeHFEvDIiPhgRf1Y69nGKozTeEBGzI2I6sJniaIl3RcTCiDgBuJbiKIrhXAscHBErIuLwiPhr4BzgqnK+YWks6WuYQ2FGCwCFGS30NczJOJEkSZKkrE3UERjfpDiC4ucUy4iVwOdLr70T+CjwD8AfAM8Ad1O6fWpK6amI+HvgU8DXgH9OKZ0TEW8FvgisAR4G/hZ4zsKeu5NSao2I15eufz7FkSDv23VRUGlXefm2fs79K5mz9vqynnNyzwbqt7YOjsCY3LOhbOfuXPB6Nhy5vGznkyRJklQd8dy1LFUpEbEAaAWOSSn9YpTnSMCfp5S+vbf91qxZ439UjWmugSFJkiRNXIsWLYrdbZ+oIzCy9P9FxNqU0itGekBEXAu8vYKZpFzZ0nwq2+a8ksnb2umb1kT/lAOzjiRJkiQpYxYY1fMk8JLS49697bgbH+P3a2K0ly2RlGP9Uw60uJAkSZI0yAKjSlJK/RTXxhjNsRuA8i0CIEmSJEnSGDNR70IiSZIkSZLGEAsMSZIkSZKUexYYkiRJkiQp91wDYxxqb2+nUChkHUOSJEmSpH22aNGi3W63wBiHCoUCU6dOzTqGJEmSJEll4xQSSZIkSZKUexYYkiRJkiQp9ywwJEmSJElS7llgSJIkSZKk3LPAkCRJkiRJuWeBIUkVtKWQeLhzgC2FlHUUSZIkaUzzNqqSVCF3tfdzw4P9zJteS0dXgbcdVsvxTf61K0mSJI2G/5Kuooi4HTih9PT4lNLPRnDMAqC19PT+lNKiyqST8u3Kewr7fY6ne3awaXsZwoxQXU2w4vR5NDfW0dbZy0WrOvjqfX0VudaBU+CghuoOqrv4mPqqXk+SJEkTm1NIqu/rQBPwS4CI+GhE3BkR2yJid2PMnyjtf3UVM0oqg9lTa2hurAOgubGOg6bWZJxIkiRJGrscgVF93SmljiHP64HvArcDl+66c0ppAOiIiK7qxJPyqRzf9n/v4T5+8Gh/GdKMzMbuAdo6ewdHYDzdPVCxa716fi1nLJxcsfNLkiRJWbPAyFhK6WMAEXFm1lmk8e6MhZOr+iH/rvZ+PnxLR2kNjH7OPsI1MCRJkqTR8l/SklQhxzfVcsQBNWzs2cHshnpm1kfWkSRJkqQxywJDkipoZn0ws961LyRJkqT95SKeklRBdb2bmbH1Qep6N2cdRZIkSRrTHIEhSRUyd/0dHProSnqnzadu2zoeevFy1s89YfgDJUmSJD2PBYakXHj5vZdlHWHQlO0baChs3O/z7Kip55GTV1KY0UL91lYOv205R/5uxf4H3IOe+tlsnzKnYucfrV8v/mTWESRJkjQOOIUkYxHRHBFLgAWl50tKP9MzDSZpv/U1zKEwowWAwowW+hryVy5IkiRJY4UjMLJ3BXD2kOe/Lv1eCtxe9TRSRvL0LX3LYzfS0nbTfp9ncs8G6re2Do7AmNyzoQzp9qxj7km0LjiroteQJEmSsmKBkbGU0jnAORnHkDRE64KzylIEzF1/B4euPn9wDYy1C893DQxJkiRplCwwqu/ciDgHWJpSume4nSOiGXgAqAMeqnA2SWW0fu4JbJ51FFO2r2f7lLn01s3KOpIkSZI0ZllgVNcyoKH0+IkRHrMOWFJ6XCh3IEmV1Vs3y+JCkiRJKgMLjCpKKT01imP6gYcrEEeSJEmSpDHDu5BIkiRJkqTcs8CQJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlnot4jkP19fV0d3dnHUOSJEmSpLKxwBiHmpqaso4gSZIkSVJZOYVEkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYYkiRJkiQp9ywwpHFsc88AazcW2NwzkHUUSZIkSdov3oVEGqdWt3Zx3T2bmfeCWjqe7efcY2axtGV61rEkSZIkaVQsMKTd+MitHWU93/qufjZ2V3cURF1NsOL0eTQ31tHW2ctFqzq4+s5NZb/O7Kk1zJ1evb9KPnPKvKpdS5IkSVJ+OIWkiiLi9ohIpZ/jRnjMgiHHrKl0Ro0fs6fW0NxYB0BzYx0HTa3JOJEkSZIkjZ4jMKrv68ClwKaIWAD8HbAUaALagZuAK1JKPaX9nyi99kHgtKqnnaDK/S3/N+/t5Ib7tpT1nMPZ2D1AW2fv4AiMpys0AuR1h0xn2eLGipxbkiRJknaywKi+7pRSB0BEHAbUAOcD/wMcDlwHHAicC5BSGgA6IqIrm7gqh2WLG6v+IX91axcX37p+cA2M9x53gGtgSJIkSRqzLDAylFL6EfCjIZsejYhPAZ+gVGBIo7W0ZTpL5jXQ0dXPvOm1zGpwCokkSZKkscsCI39mAJuzDqHxYVZDjcWFJEmSpHHBRTxzJCJeRHGtiy9lnUWSJEmSpDyxwMiJiJhLcTrJfwKfzziOJrja7Zto2LSG2u3lv+2qJEmSJI2GU0hyICLmAT8B1gDvSCmljCNpApvZdgtNv1lB77SDqdv2FO1LLmJL86lZx5IkSZI0wVlgZCwimoDVwP3A21JK/RlHUknL7RdmHWG3Jm9rp66no2Ln31FTzyMnr6Qwo4X6ra0ccttyXnj3FWU5d2/DPPqmNZXlXPuq9cRrMrmuJEmSpPJwCkmGImI+cAfQAVwEHBQR80o/rryoTPQ1zKEwowWAwowW+hrmZJxIkiRJkhyBkbVTgJeUftp2ea0FeKzagfR7ef3Gfs79K5mz9vqKnX9yzwbqt7YOjsCY3LOhbOfuXPB6Nhy5vGznkyRJkjRxWGBkKKX0DeAbGcfQGLPhyOUVLQFmtt1Cy+0XDK6B8dQrLnYNDEmSJEmZs8CovnMj4hxgaUrpnuF2johm4AGgDniowtkktjSfyrY5r2Tytnb6pjXRP+XArCNJkiRJkgVGlS0DGkqPnxjhMeuAJaXHhXIHknanf8qBFheSJEmScsUCo4pSSk+N4ph+4OEKxJEkSZIkaczwLiSSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTccw2Mcai9vZ1CwfU+JUmSJEljz6JFi3a73QJjHCoUCkydOjXrGJIkSZIklY1TSCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkjcCWQuLhzgG2FFLWUSRJkqQJyduoStIw7mrv54YH+5k3vZaOrgJvO6yW45v861OSJEmqJv8FXkURcTtwQunp8Smln43gmAVAa+np/SmlRZVJJ1XelfcUKnLep3t2sGl7RU4NQF1NsOL0eTQ31tHW2ctFqzr46n19Zb/OgVPgoIbqD4y7+Jj6ql9TkiRJ2ldOIam+rwNNwC8jYlJE/CAi2iJie0S0R8S/RsTBQ/Z/orT/1ZmklcTsqTU0N9YB0NxYx0FTazJOJEmSJE08jsCovu6UUgdAREwCfgJ8GmgHDgauAv4vcCxASmkA6IiIrmziSuVTqW/6v/dwHz94tL8i5wbY2D1AW2fv4AiMp7sHKnKdV8+v5YyFkytybkmSJGmss8DIUEppB7BiyKbHI+IzwPcjYkpKqYKD4qXx44yFkyv6wf+u9n4+fEtHaQ2Mfs4+wjUwJEmSpGrzX+A5EhEHAMuAn1teSPlxfFMtRxxQw8aeHcxuqGdmfWQdSZIkSZpwXAMjByLiyojYBmwCmoE3ZhxJ0i5m1gcLG2ssLyRJkqSMWGDkw2eBlwOnAAPAv0aEn5KkHKvr3cyMrQ9S17s56yiSJEnShOAUkhxIKT0NPA08FBFrKd555DXAf2UaTNJuzV1/B4c+upLeafOp27aOh168nPVzTxj+QEmSJEmjZoGRPztHxVTmdg1SBb383suyjrBXU7ZvoKGwcb/Ps6OmnkdOXklhRgv1W1s5/LblHPm7FfsfcBc99bPZPmVO2c87Gr9e/MmsI0iSJGmCs8DIUEQcDxwN/BToBA4BPgE8VtomKYf6GuZQmNECQGFGC30Nc6jveiLjVJIkSdL4ZoGRrR7gTOAKYBrQDvwIeKt3IdFYlPdv6Vseu5GWtpv2+zyTezZQv7V1cATG5J4NZUj3fB1zT6J1wVkVObckSZI01lhgZCil9BtgadY5pImidcFZZSkE5q6/g0NXnz+4Bsbahee7BoYkSZJUYRYY1XduRJwDLE0p3TPczhHRDDwA1AEPVTibpBFYP/cENs86iinb17N9ylx662ZlHUmSJEka9ywwqmsZ0FB6PNIJ8+uAJaXHhXIHkjQ6vXWzLC4kSZKkKrLAqKKU0lOjOKYfeLgCcSRJkiRJGjMmDb+LJEmSJElStiwwJEmSJElS7llgSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs8CQ5IkSZIk5Z4FhiRJkiRJyj0LDEmSJEmSlHsWGJIkSZIkKfcsMCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTcs8CQJEmSJEm5Z4EhSZIkSZJyL5cFRkScGBEpIg7KOsvejJWckiRJkiSNdbksMID/BpqATVkHqYaIuD0irsk6h6T8qevdzIytD1LXuznrKJIkSVKmarMOsDsppV6gI+sckpSluevv4NBHV9I7bT5129bx0IuXs37uCVnHkiRJkjKRaYEREX8E/AOwCBgAfgf8FXAQsBqYnVJ6urTvXwEfL712G7AK+MeUUpRe/zhwJvBZ4HJgNvAt4N3AXwOXAFOBfwI+mFLaUTru7cD7gcOAHuAO4KKU0lP78FaOi4hPls5xP3BuSumXQ97nq4D/DRwDbAZ+AFycUtoaEd8ATgBOiIj3lA5pAZ4ArgNOAuYBTwJfBa7amV1S+bz83ssqdu4p2zfQUNi4z8ftqKnnkZNXUpjRQv3WVg6/bTlH/m5FWbP11M9m+5Q5ZT3ncH69+JNVvZ4kSZLGh8ymkERELfB94KfAYuAPgRUUi4xd9z0e+Brwj8ASigXA5bs57QLgT4A3An8G/Hlp32OAUygWGe8F/nTIMXXA35cyvJFiQXLDPr6dq4CLgVcCjwI3R8TUUvaXAbeWciwu5VoCXF869v3AXcDXKU6baaJYXkwCngLeAhwOfBS4FHjnPmaTNEb1NcyhMKMFgMKMFvoaqls0SJIkSXmS5QiMGUAj8B8ppUdK2x4EiIi5u+z7PuDWlNKVpecPRcQxwLt22a8GeGdKaQuwJiJ+RHF0w8GlaSlrI+JOYCnwHYCU0vVDjn80Is4v7fcHKaUnR/hePpFSuqWU/Z0UR0v8BcXS5UPATSmlq3fuXLrGryNiTkppQ0T0At0ppaHTZgaAjw15/lhEHA28DVg5wlySRqiSowJaHruRlrab9vm4yT0bqN/aOjgCY3LPhrJn65h7Eq0Lzir7eSVJkqRyy6zASCk9U5o+cUtE3EZxWsi3U0ptu9n9MOA/dtn2c55fYLSVyoud1gMPlcqLodsGv8YslQJ/T3FUxAFAlF5qplhEjMRdOx+klLoi4j7giNKmVwALI+KtQ/bfeY1DgD1+IomI8yiOGnkR0ABMBh4fYSZJOdG64KxRlQRz19/BoavPH1wDY+3C810DQ5IkSRNWpmtgpJTeGRErgNOANwOfiogzgMIoT9m36yX2sK0GICKmAbcAPwbeQbFMOAj4L4pTS8phEsWRGJ/fzWt7XGejVHisAD5I8a4sW4H38NzpL5LGsfVzT2DzrKOYsn0926fMpbduVtaRJEmSpMxkfheSlNK9wL3AlRGxCjib4uKVQz1IcR2LoY4tw+UPo1hYXJpSagWIiD8bxXmOo7j2xc5SZBHwz6XXfgUcmVJ6eC/H91IqVYZ4DfDzlNLg7VUj4pBRZJM0hvXWzbK4kCRJksh2Ec+WiPhMRLwqIl4UEUuBo4AHdrP7F4FTIuJDEfGSiFhOeUYitFEc7XFhRLw4It4AfGIU57ksIv44Io6kuDhnL/BvpdeuBI6NiGsj4uURsTAi3hgRXxly/GOlfRZExEERMQl4CDg6Ik4vvee/o7iehyRJkiRJE05mBQbQDRwK/DvFD+v/BHyT4gf+50gp3UVxvYv3Ab8Fzijtt31/AqSUNlIc8XEGxeLk74EPjOJUHwGupjja4iXAG1NK20rX+C3wRxTvkHIHxdEm/5viWhw7XUWx9HgA2Ehx/Y2vULwN7L8B95SOvxpJkiRJkiagSCllnWFUIuLzwOtSSi/LOkve3HzzzWnq1KlZx5AkSZIkaZ+ddNJJsbvtma+BMVIR8SHgP4Eu4HXAecClmYaSJEmSJElVkeUUkn31Sop3DFkDvB+4hOJdOiqmtG5F1x5+rq3ktSVJkiRJ0u+NmREYKaW3ZnDZj1Fcn2J3tlYziCRJkiRJE9mYKTCykFLaAGzIOockSZIkSRPdWJpCIkmSJEmSJigLDEmSJEmSlHsWGJIkSZIkKfcsMCRJkiRJUu65iOc4VF9fT3d3d9YxJEmSJEkqGwuMcaipqSnrCJIkSZIklZVTSCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTcs8CQlGubewZYu7HA5p6BrKNIkiRJylBt1gEkaU9Wt3Zx3T2bmfeCWjqe7efcY2axtGV61rEkSZIkZcACQ5rAPnJrR0XPv76rn43dox85UVcTrDh9Hs2NdbR19nLRqg6uvnPTbvedPbWGudNH/lfaZ06ZN+pckiRJkqrPKSRVFBHnREQq/Vy7j8c+NuTYgyqVUcqT2VNraG6sA6C5sY6DptZknEiSJElSVhyBUX3dwCGl34Mi4gLgQ0ATcD9wUUrpv4bscgzwWuA7VcqpCaDSoxC+eW8nN9y3ZdTHb+weoK2zd3AExtN7Gc3xukOms2xx46ivJUmSJCnfLDCqL6WUnjNuPyLeCnwBuAD4aen3qog4IqXUVjpoY0Q8U/W00n5Ytrhxv0qF1a1dXHzr+sE1MN573AGugSFJkiRNULmeQhIR0yLinyOiKyLWR8QlEXFzRHyj9PrbI+KeiHg2IjZExL9HxMFDjj+xNOXi9Ij4ZUT0RMR/RcQfRMQJEXFv6dw3R8SBQ477RmnbxRHRERFbIuIzETEpIj5eulZHRFy8S94PRMRvI2JbRDwVEV+LiMYRvNUPAN9IKX01pbQ2pfReoB04vxx/jtJYtbRlOl9603zOfeUBfOlN8y0vJEmSpAks1wUGcDVwAvCnwEnAYorTKHaqA/6+tP2NwEHADbs5z+XARcAfArOAm4CPAecCJwJHAh/f5Zg/AlpKr58HfBj4f0A98JrS/p+JiFcMOWZH6TpHAn8BHAv8n729wYioA14B3LrLS7cCr9rbsdJEMKuhhsNn1zOrwfUvJEmSpIkst1NIImI68FfAX6aU/rO0bTnw5M59UkrXDznk0Yg4H1gbEX+QUnpyyGt/t3M9idLimf8HeEVK6Velbf8EnLlLhC3Ae1JKA8CDEfG3QFNK6bTS6w9FxEeApcAvS3lWDDn+sYj4MPD9iDg7pbRjD2/1IKAGWL/L9vXA6/ZwjCRJkiRJE0qeR2AcAkwG7t65IaW0DViz83lEHB0R34+IxyPiWeAXpZeadznXb4c83lkU3LfLtjm7HPNAqbwYus+aXfZ5znERcVJE/GdEPFnK812Ko0S8X6O0j2q3b6Jh0xpqt+/+tqmSJEmSJpY8Fxh7FRHTgFso3s3jHRTv0rFzdETdLrv3DXmcAFJKu27b9c+ib5fnaQ/bJpXyvAj4IbAW+HOK00L+ag95hnoaGADm7rJ9LtDx/N2l8W9m2y0svPXtNP1mBQtvfTsz227JOpIkSZKkjOV2CgnwCMXC4BjgUYCImAosKr12GMXpF5emlFpLr/9ZNlEBeCXFouJvdo7ciIg3DndQSqk3In4J/DHw70Ne+mO8ZarKrOX2Cyt+jcnb2qnr2b/ubUdNPY+cvJLCjBbqt7ZyyG3LeeHdV+z1mN6GefRNaxr1NVtPvGbUx0qSJEmqvNyOwEgpdQHXA1dGxMkRcQTwNYqZE9AGFIALI+LFEfEG4BOZBYb/KWW7KCJaIuJtFBf0HInPAedExF9HxOER8QVgPnBtZaJK+dbXMIfCjBYACjNa6GvYdYaXJEmSpIkmzyMwAD4ITAN+AHQBn6c4tWJ7SmljRJwNfBp4D8V1Lj4A/CiLoCml30bE+4GLgU8C/00x/00jOPam0m1cLwOaKK618fqU0uMVjKwJqBqjDObcv5I5a68ffse9mNyzgfqtrYMjMCb3bBj2mM4Fr2fDkcv367qSJEmS8itSSllnGLGIqAceBz6bUro66zz7KiLOAa5JKU0f5fEnAquB2Smlp/e035o1a8bOf1RpN2a23ULTb1bQO+1g6rY9RfuSi9jSfGrWsSRJkiRVwaJFi2J323M9AiMiXg4cTvFOJC+gOLrhBYxgVEOOTYuILuBrKaWLRnpQRNwPvLhiqaQc2dJ8KtvmvJLJ29rpm9ZE/5QDs44kSZIkKWO5LjBKPgC8FOgHfgP8UUrpyUwTjd53gJ+WHm/Zx2NfT/G2sgDPlC2RlFP9Uw60uJAkSZI0KNcFRkrp1xTv7jEupJSeBZ4d5bGuhyFJkiRJmrByexcSSZIkSZKknSwwJEmSJElS7llgSJIkSZKk3Mv1Ghganfb2dgqFQtYxJEmSJEnaZ4sWLdrtdguMcahQKDB16tSsY0iSJEmSVDZOIZEkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSxpUthcTDnQNsKaSso0iSJEkqI2+jKmncuKu9nxse7Gfe9Fo6ugq87bBajm/yrzlJkiRpPPBf9lUUEbcDJ5SeHp9S+tkIjlkAtJae3p9SWlSZdFJlXXlPoeznfLpnB5u2//55XU2w4vR5NDfW0dbZy0WrOvjqfX17PP7AKXBQQ3kHol18TH1ZzydJkiSpyCkk1fd1oAn4JUBEfDQi7oyIbRGxuzHvT5T2v7qKGaUxafbUGpob6wBobqzjoKk1GSeSJEmSVC6OwKi+7pRSx5Dn9cB3gduBS3fdOaU0AHRERFd14kmVUYmRCd97uI8fPNo/+Hxj9wBtnb2DIzCe7h7Y6/Gvnl/LGQsnlz2XJEmSpPKzwMhYSuljABFxZtZZpLHmjIWTn1NA3NXez4dv6SitgdHP2Ue4BoYkSZI0Xvgve0njxvFNtRxxQA0be3Ywu6GemfWRdSRJkiRJZWKBIWlcmVkfzKx37QtJkiRpvHERT0njSl3vZmZsfZC63s1ZR5EkSZJURo7AkDRuzF1/B4c+upLeafOp27aOh168nPVzTxj+QEmSJEm5Z4EhqWJefu9lFTv3lO0baChsfM62HTX1PHLySgozWqjf2srhty3nyN+t2Ot5eupns33KnLLn+/XiT5b9nJIkSdJE5hSSjEVEc0QsARaUni8p/UzPNJg0BvU1zKEwowWAwowW+hrKX0xIkiRJyoYjMLJ3BXD2kOe/Lv1eCtxe9TRSGVVyFELLYzfS0nbTc7ZN7tlA/dbWwREYk3s2DHuejrkn0brgrErFlCRJklQmFhgZSymdA5yTcQxpzGldcNbzioe56+/g0NXnD66BsXbh+a6BIUmSJI0TFhjVd25EnAMsTSndM9zOEdEMPADUAQ9VOJs0pq2fewKbZx3FlO3r2T5lLr11s7KOJEmSJKlMLDCqaxnQUHr8xAiPWQcsKT0ulDuQNN701s2yuJAkSZLGIQuMKkopPTWKY/qBhysQR5IkSZKkMcO7kEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWei3iOQ/X19XR3d2cdQ5IkSZKksrHAGIeampqyjiBJkiRJUlk5hUSSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3LDAkjSubewZYu7HA5p6BrKNIkiRJKiPvQiJp3Fjd2sV192xm3gtq6Xi2n3OPmcXSlulZx5IkSZJUBhYYkqrqI7d2lO1c67v62dj9+5EWdTXBitPn0dxYR1tnLxet6uDqOzcNe57ZU2uYO728fx1+5pR5ZT2fJEmSNNE5haSKIuL2iEiln+NGeMyCIcesqXRGaSybPbWG5sY6AJob6zhoak3GiSRJkiSViyMwqu/rwKXApohYAPwdsBRoAtqBm4ArUko9pf2fKL32QeC0qqeVyqycIxO+eW8nN9y3ZfD5xu4B2jp7B0dgPN09snUwXnfIdJYtbixbLkmSJEnlZ4FRfd0ppQ6AiDgMqAHOB/4HOBy4DjgQOBcgpTQAdEREVzZxpfxatrjxOcXD6tYuLr51/eAaGO897gDXwJAkSZLGCQuMDKWUfgT8aMimRyPiU8AnKBUYkkZuact0lsxroKOrn3nTa5nV4BQSSZIkabywwMifGcDmrENIY9WshhqLC0mSJGkcchHPHImIF1Fc6+JLWWeRJEmSJClPLDByIiLmUpxO8p/A5zOOI41ptds30bBpDbXbh7+FqiRJkqSxwSkkORAR84CfAGuAd6SUUsaRpDFrZtstNP1mBb3TDqZu21O0L7mILc2nZh1LkiRJ0n6ywMhYRDQBq4H7gbellPozjiRVTMvtF5b1fJO3tVPX0/GcbTtq6nnk5JUUZrRQv7WVQ25bzgvvvmLYc/U2zKNvWlNZ8+2q9cRrKnp+SZIkaTxzCkmGImI+cAfQAVwEHBQR80o/rkIojUJfwxwKM1oAKMxooa9hTsaJJEmSJJWDIzCydQrwktJP2y6vtQCPVTuQVEnlHoEw5/6VzFl7/XO2Te7ZQP3W1sERGJN7NozoXJ0LXs+GI5eXNZ8kSZKk8rHAyFBK6RvANzKOIY1ZG45c/rzSYWbbLbTcfsHgGhhPveJi18CQJEmSxgELjOo7NyLOAZamlO4ZbueIaAYeAOqAhyqcTRrztjSfyrY5r2Tytnb6pjXRP+XArCNJkiRJKgMLjOpaBjSUHj8xwmPWAUtKjwvlDiSNR/1TDrS4kCRJksYZC4wqSik9NYpj+oGHKxBHkiRJkqQxw7uQSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs81MMah9vZ2CgXX+5QkSZIkjT2LFi3a7XYLjHGoUCgwderUrGNIkiRJklQ2TiGRJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYYksaFLYXEw50DbCmkrKNIkiRJqgBvoyppzLurvZ8bHuxn3vRaOroKvO2wWo5v8q83SZIkaTzxX/hVFBG3AyeUnh6fUvrZCI5ZALSWnt6fUlpUmXRS+Vx5T6Fi5366Zwebtj93W11NsOL0eTQ31tHW2ctFqzr46n19zzv2wClwUEP5Bp5dfEx92c4lSZIkae+cQlJ9XweagF9GxKSI+EFEtEXE9ohoj4h/jYiDh+z/RGn/qzNJK40Bs6fW0NxYB0BzYx0HTa3JOJEkSZKkcnMERvV1p5Q6ACJiEvAT4NNAO3AwcBXwf4FjAVJKA0BHRHRlE1fad5UcmfC9h/v4waP9z9m2sXuAts7ewREYT3cP7PbYV8+v5YyFkyuWTZIkSVLlWGBkKKW0A1gxZNPjEfEZ4PsRMSWltH33R0oT1xkLJz+vhLirvZ8P39JRWgOjn7OPcA0MSZIkabzxX/g5EhEHAMuAn1teSCN3fFMtRxxQw8aeHcxuqGdmfWQdSZIkSVKZuQZGDkTElRGxDdgENANvzDiSNObMrA8WNtZYXkiSJEnjlAVGPnwWeDlwCjAA/GtE+ClM2kd1vZuZsfVB6no3Zx1FkiRJUpk5hSQHUkpPA08DD0XEWop3HnkN8F+ZBpPGkLnr7+DQR1fSO20+ddvW8dCLl7N+7gnDHyhJkiRpTLDAyJ+do2IqdxsHqUpefu9lZT3flO0baChs3O1rO2rqeeTklRRmtFC/tZXDb1vOkb9bsdt9e+pns33KnLJm2+nXiz9ZkfNKkiRJE50FRoYi4njgaOCnQCdwCPAJ4LHSNkkj1Ncwh8KMFgAKM1roa5hDfdcTGaeSJEmSVC4WGNnqAc4ErgCmAe3Aj4C3ehcSjQflHo3Q8tiNtLTdtNvXJvdsoH5r6+AIjMk9G/Z4no65J9G64KyyZpMkSZJUWRYYGUop/QZYmnUOaaxoXXDWHouHuevv4NDV5w+ugbF24fmugSFJkiSNIxYY1XduRJwDLE0p3TPczhHRDDwA1AEPVTibNGatn3sCm2cdxZTt69k+ZS69dbOyjiRJkiSpjCwwqmsZ0FB6PNLJ+euAJaXHhXIHksaT3rpZFheSJEnSOGWBUUUppadGcUw/8HAF4kiSJEmSNGZMGn4XSZIkSZKkbFlgSJIkSZKk3LPAkCRJkiRJuWeBIUmSJEmScs9FPMeh+vp6uru7s44hSZIkSVLZWGCMQ01NTVlHkCRJkiSprJxCIkmSJEmScs8CQ5IkSZIk5Z4FhiRJkiRJyj0LDEmSJEmSlHsWGJLGhc09A6zdWGBzz0DWUSRJkiRVgHchkTTmrW7t4rp7NjPvBbV0PNvPucfMYmnL9KxjSZIkSSojCwxJZfWRWzvKer71Xf1s7N77qIq6mmDF6fNobqyjrbOXi1Z1cPWdmwZfnz21hrnTy/vX3WdOmVfW80mSJEnaO6eQVFFE3B4RqfRz3AiPWTDkmDWVziiNRbOn1tDcWAdAc2MdB02tyTiRJEmSpHJzBEb1fR24FNg0dGNETAF+DhwFHJNS+kXppSeAJuCDwGlVzCmNSrlHJnzz3k5uuG/LXvfZ2D1AW2fv4AiMp3cZsfG6Q6azbHFjWXNJkiRJqi4LjOrrTintboz9VcCTFAuMQSmlAaAjIrqqEU7Km2WLG4ctH1a3dnHxresH18B473EHuAaGJEmSNM5YYORARPwJsBQ4E3h9xnGkMWdpy3SWzGugo6ufedNrmdXgFBJJkiRpvLHAyFhE/AHwZeB0oCfjONKYNauhxuJCkiRJGsdcxDNDEVEDfBO4OqV0b9Z5JEmSJEnKKwuMbF0K9AKfyzqINJbVbt9Ew6Y11G7fNPzOkiRJksYkp5Bk62TgtUBfRAzd/rOIuCmltCybWNLYMbPtFpp+s4LeaQdTt+0p2pdcxJbmU7OOJUmSJKnMLDCy9U5g2pDn84FbgGXAnZkkkvZDy+0XlvV8k7e1U9ezu5v2/N6OmnoeOXklhRkt1G9t5ZDblvPCu694zj69DfPom9ZUtlytJ15TtnNJkiRJGhkLjAyllFqHPh9yq9RHUkpPZhBJGnP6GuZQmNECQGFGC30Nc6jveiLjVJIkSZLKzQJDUtmUe2TCnPtXMmft9XvdZ3LPBuq3tg6OwJjcs+F5+3QueD0bjlxe1mySJEmSqssCI0dSSo8BMdx+0kSx4cjlwxYPM9tuoeX2CwbXwHjqFRe7BoYkSZI0DllgVN+5EXEOsDSldM9wO0dEM/AAUAc8VOFs0pizpflUts15JZO3tdM3rYn+KQdmHUmSJElSBVhgVNcyoKH0eKST9NcBS0qPC+UOJI0H/VMOtLiQJEmSxjkLjCpKKT01imP6gYcrEEeSJEmSpDFjUtYBJEmSJEmShmOBIUmSJEmScs8CQ5IkSZIk5Z5rYIxD7e3tFAqu9ylJkiRJGnsWLVq02+0WGONQoVBg6tSpWceQJEmSJKlsnEIiSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKPQsMSZIkSZKUexYYksa1LYXEw50DbCmkrKNIkiRJ2g/ehUTSuHVXez83PNjPvOm1dHQVeNthtRzf5F97kiRJ0ljkv+Ql5cKV9xTKcp6ne3awaXvxcV1NsOL0eTQ31tHW2ctFqzr46n19+3zOA6fAQQ2VGbB28TH1FTmvJEmSNN44haSKIuL2iEiln+NGeMyCIcesqXRGaTyZPbWG5sY6AJob6zhoak3GiSRJkiSNliMwqu/rwKXAJoCIeAx40S77XJlS+kjp8RNAE/BB4LQqZZSqrlwjEb73cB8/eLQfgI3dA7R19g6OwHi6e2BU53z1/FrOWDi5LPkkSZIkjY4FRvV1p5Q6dtl2BfDlIc+7dj5IKQ0AHRHRhaRhnbFw8mDZcFd7Px++paO0BkY/Zx/hGhiSJEnSWOW/5PPh2d2UGpL20/FNtRxxQA0be3Ywu6GemfWRdSRJkiRJo+QaGPnwwYjYFBG/iYiPRkRd1oGk8WJmfbCwscbyQpIkSRrjHIGRvS8Cv6a4JsaxwGeAFuCvswwlSZIkSVKeWGBkLKX0uSFPfxsRW4GbIuLilNKmrHJJ40ld72ambF/P9ilz6a2blXUcSZIkSaNggZE/Py/9XkjpTiWSRm/u+js49NGV9E6bT922dTz04uWsn3tC1rEkSZIk7SMLjPxZUvrdnmUIqVpefu9lZTnPlO0baChsfN72HTX1PHLySgozWqjf2srhty3nyN+t2Kdz99TPZvuUOWXJuTu/XvzJip1bkiRJGi8sMDIUEccDxwGrgS3AMcDngR+klNqyzCaNF30NcyjMaAGgMKOFvoY51Hc9kXEqSZIkSfvKAiNbBeCtwN8D9cDjwFeBf8gylFRN5Rp90PLYjbS03fS87ZN7NlC/tXVwBMbkng37fO6OuSfRuuCscsSUJEmSNEoWGBlKKf2K4ggMSfupdcFZuy0Z5q6/g0NXnz+4Bsbahee7BoYkSZI0BllgVN+5EXEOsDSldM9wO0dEM/AAUAc8VOFs0rizfu4JbJ51lHchkSRJksY4C4zqWgY0lB6PdBL+On6/sGeh3IGkiaC3bpbFhSRJkjTGWWBUUUrpqVEc0w88XIE4kiRJkiSNGZOyDiBJkiRJkjQcCwxJkiRJkpR7FhiSJEmSJCn3LDAkSZIkSVLuWWBIkiRJkqTcs8CQJEmSJEm5Z4EhSZIkSZJyzwJDkiRJkiTlngWGJEmSJEnKvdqsA6j86uvr6e7uzjqGJEmSJEllY4ExDjU1NWUdQZIkSZKksnIKiSRJkiRJyj0LDEmSJEmSlHsWGJIkSZIkKfcsMCRJkiRJUu5ZYEiSJEmSpNyzwJAkSZIkSblngSFJkiRJknLPAkOSJEmSJOWeBYYkSZIkSco9CwxJkiRJkpR7FhiSJEmSJCn3LDAkSdKobO4ZYO3GApt7BrKOIkmSJoDarANIkqSxZ3VrF9fds5l5L6il49l+zj1mFktbpmcdS5IkjWMWGJKk3PrIrR1ZR8jE+q5+Nnbne1RDXU2w4vR5NDfW0dbZy0WrOrj6zk1Zxxq12VNrmDs93/8s+swp87KOIElSpjKZQhIRJ0ZEioiDsrh+OUXExyNizT4ec3tEXFOpTJIkVdrsqTU0N9YB0NxYx0FTazJOJEmSxrusvmr4b6AJGLtf1UiSKm6ifuP8zXs7ueG+LVnH2KuN3QO0dfYOjsB4OucjRobzukOms2xxY9YxJEnSXmRSYKSUeoGJOS5YkqRhLFvcmPsP06tbu7j41vWDa2C897gDXANDkiRVVEWnkETEH0XEzyKiKyK2RMTdEbFod1NIIuKvIqItIroj4j8i4oKISENe/3hErImIsyPisYjYFhFfj4i60r5PRMSmiPhcREwactzbI+KeiHg2IjZExL9HxMEjzF8TESsjojUieiLifyLiw0PPv5tjvhERN0fEZRGxvvTevx4RDbvsOikiPh0RT5dyXVWu3JIkVdrSlul86U3zOfeVB/ClN823vJAkSRVXsREYEVELfB9YCSwDJgNHA88bYxoRxwNfAy4B/i9wAvDp3Zx2AfAnwBuBg4HvUJyK0g6cAhwGfAu4s/QaQB3w98CDwEHAlcANwB+N4G1MAp4C3gJsBI4FrqM49WXlXo47AegBTi7lvL503fcN2WcZ8AXgVcAS4N+AX5ay7W9uSZIqblZDDbMaXPtCkiRVR6SUht9rNCeOOIDiB/0TU0p37PLaicBqYHZK6emIuAGYlVI6bcg+1wHvSilF6fnHgYuBeSmlLaVt36ZYFhxcmpZCRNwOrEkpXbiHXIcBa4EXppSeHMX7+gzwypTS64bkOjOltKj0/BvAGcAfpJS6StveTrHwOCCltK2UsT6ldPyQ8/4n8HhK6a/3N/eaNWsq8x9VkiRJkqQKW7RoUexue8WmkKSUngG+AdwSET+MiA9ERPMedj8MuHuXbT/fzX5tO8uLkvXAQzvLiyHb5ux8EhFHR8T3I+LxiHgW+EXppT1leY6IOC8ifhERGyOiC/ibERz7253lRcldFEdUHDJ0n12OWVfO3JIk7Yva7Zto2LSG2u2ury1JkvKpomtgpJTeCfwh8P8BbwZ+FxGn7scp+3a9xB62TQKIiGnALUA38A7gGGDnKI+64S4WEW8FVlAsYk6lONXjSyM5dgQqlluSpH0xs+0WFt76dpp+s4KFt76dmW23ZB1JkiTpeSp+F5KU0r3AvcCVEbEKOJviOhJDPUjxQ/pQx5bh8odRXD/i0pRSK0BE/Nk+HP8a4OcppWt2boiIQ/ay/04vi4hpKaVtpefHAb3AIyO87v7mlqRxqeX23c4OHHMmb2unric/N+PaUVPPIyevpDCjhfqtrRxy23JeePcVWcfab70N8+ib1pR1jFxoPfGa4XeSJCnnKjYCIyJaIuIzEfGqiHhRRCwFjgIe2M3uXwROiYgPRcRLImI58KdliNEGFIALI+LFEfEG4BP7cPz/3979h9pd13Ecf778/SsXOaYrGCr+qMxQW1Bp/ghFa/1hIChRaNlkqP0wypgmzhARKnVkoUbmyECxQJLwR5IW+Cu1ZG2aGVqWmXPoNrZ0Tnn3x/dgl+t2z7lu937O2Z4PuH+c7zn33he8ufec8z6fz+f9V+CIJJ/s5bqI7syNfnYArk9ySJITgMuBH49paEx1bkmSBrZh11ms33M/ANbvuR8bdp3V5zskSZKm31SuwPgvcBBwC91qgheAn9NN0zhy7AOr6oEk84FLgO8Ad/ced+nmBKiqF5OcTjfR5By6cye+Dtwx4I+4lv9PCAndZJPvA1/s832/A5bTHVS6W+/7zp/G3JK0VdpaPkWetfwnzHri+tYx3rTjKyvYec0zb67A2PGVFa0jbRGr9v0UKw45s3UMSZK0hUzZFJLNleRK4PiqOrR1lsnoTSGZWVWfbpXBKSSSpMmY8eydzH7sKl7b/T3stO45nj/sa6yeszlHVkmSJL19m5pCMuVnYAwqyTeB3wBrgeOBBcAFTUNJkrQNWD3nRNbNmsuO655nw+6zeX2XvVpHkiRJeouhaWAAc4FvADOAZ4CFwOKp/IVJrgE+t4m7b6yqBVP5+yVJGhav77KXjQtJkjTUhnYLyXRIMgvYcxN3r6mqkdwE7BYSSZIkSdKoGvotJC30GhQj2aSQJEmSJGlbMmVjVCVJkiRJkrYUGxiSJEmSJGno2cCQJEmSJElDb5s+xHNrleQOYGbrHCNuJrCydQhNmnUbXdZuNFm30WXtRpe1G03WbXRZuzZWVtVJ4y/awJA2IskjVTW3dQ5NjnUbXdZuNFm30WXtRpe1G03WbXRZu+HiFhJJkiRJkjT0bGBIkiRJkqShZwND2rjrWgfQ22LdRpe1G03WbXRZu9Fl7UaTdRtd1m6IeAaGJEmSJEkaeq7AkCRJkiRJQ88GhiRJkiRJGno2MCRJkiRJ0tCzgSFNIMlZSe5JsipJJdm3dSZtXJKzkzyT5NUkjyb5eOtMmliSo5P8Kslzvb+vM1pnUn9JFiZ5OMmaJC8muS3JB1rnUn9JzkmytFe7NUkeSDKvdS5NTu9vsJJc3TqLJpZkUa9WY7/+0zqX+ksyO8mS3vPcq0keT3JM61yygSH1sxtwF7CocQ5NIMmpwGLgMuBw4H7g9iRzmgZTP3sAy4CvAq80zqLBHQv8CPgY8AngdeDuJO9qGUoD+RfwLeAIYC7wW+DWJB9smkoDS/IR4CxgaessGtiTwOwxX4e2jaN+krwTuA8IMA94H/BlYEXDWOpxCok0gCRzgYeB/arq743jaJwkDwFLq2r+mGtPAb+oqoXtkmlQSdYC51bVDa2zaHKS7AGsBk6uqtta59HkJHkJWFhV17bOooklmQH8EfgScDGwrKrObZtKE0myCDilqlylNkKSXAYcU1VHts6it3IFhqSRlmQn4EN0K2XGuovuE2JJU+sddK8nXm4dRINLsn2S0+hWQt3fOo8Gch1dY/6e1kE0Kfsn+Xdvm+tNSfZvHUh9nQw8lOTmJCuSPJbk3CRpHUw2MCSNvpnA9sAL466/AOwz/XGkbc5i4DHggcY5NIAkh/ZWPK0HrgE+U1V/bhxLfSSZDxwAfLt1Fk3KQ8AZwEnAfLrXJfcn2atlKPW1P3A28DRwIt3z3OXAOS1DqbND6wDSdEtyKXBhn4cdV1X3TkMcSRpZSa4AjgKOqqo3WufRQJ4EDgNmAKcAS5IcW1XLmqbSJiU5mO6Mp6OqakPrPBpcVd0+9naSB+neFJ8OXNEklAaxHfDImG3If0pyIF0Dw8NzG7OBoW3RVcCNfR7z7DTk0JaxEngD2Hvc9b0BT/qWpkiSK4HT6Bq+T7fOo8FU1WvA33o3H03yYeA84Mx2qdTHR+lWGy4fs4J9e+DoJAuA3atqfatwGlxVrU2yHDiwdRZN6Hng8XHXnqA7dFyN2cDQNqeqVtK96dVWoKpeS/IocAJwy5i7TgB+2SaVtHVLshg4la558ZfWebRZtgN2bh1CE7oVeGTctZ8CT9GtzHhtugPp7UmyC/BewHNMhtt9wMHjrh0E/KNBFo1jA0OaQJJ96PYrHtS79P7eaKVnq+qlZsE03hXAz5L8ge5JZwHwbrr93RpSvekVB/RubgfMSXIY8FJVuQpqSCX5IfB5ukPOXu79nwRYW1VrmwVTX0kuB34N/JPu8NXP0o3FndcwlvqoqlXAqrHXkqyj+1/p1p8hluR7wG10K3tnARcBuwNLWuZSX1fSnVVyIXAzcDjwFeCCpqkEOEZVmlBv/NXFG7nrC457HC5JzgbOp5uxvgw4r6p+3zaVJpLkWDb+KdSSqjpjWsNoYEk29cLhkqpaNJ1ZNDlJbgCOo2vMrwaWAt+tqjtb5tLkJbkXx6gOvSQ3AUfTbQF6EXgQuKiqxm9P0JBJMo9uhdPBdA2oq4EflG+em7OBIUmSJEmShp5jVCVJkiRJ0tCzgSFJkiRJkoaeDQxJkiRJkjT0bGBIkiRJkqShZwNDkiRJkiQNPRsYkiRJkiRp6NnAkCRJkiRJQ88GhiRJkiRJGnr/A2u9Tc1CiUTQAAAAAElFTkSuQmCC\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "az.plot_forest([model_13_4, model_13_6], combined=True, figsize=(17,12), hdi_prob=0.89,\n", " model_names = [\"model_13_4\", \"model_13_6\"])\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "2c95abbe", "metadata": {}, "source": [ "## 13.4 Divergent Transitions and non-centered priors" ] }, { "cell_type": "markdown", "id": "1a82b1f4", "metadata": {}, "source": [ "### R Code 13.26" ] }, { "cell_type": "code", "execution_count": 49, "id": "cb1eebc6", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mMessages from \u001b[0m\u001b[36;1mstanc\u001b[0m\u001b[36m:\u001b[0m\n", "Warning: The parameter v has 2 priors.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 50% (4000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 75% (6000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 3e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.03 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 3e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.03 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 2e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.02 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 3e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.03 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " parameters {\n", " real v;\n", " real x;\n", " }\n", " \n", " model {\n", " v ~ normal(0, 3);\n", " x ~ normal(0, exp(v));\n", " }\n", "\"\"\"\n", "\n", "posteriori = stan.build(model)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 50, "id": "158f472e", "metadata": {}, "outputs": [], "source": [ "model_13_5 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", ")" ] }, { "cell_type": "code", "execution_count": 51, "id": "b684be8b", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
v3.5151.5081.2435.4860.2080.14837.026.01.1
x-34.371286.623-142.353209.82120.34014.404290.0232.01.1
\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "v 3.515 1.508 1.243 5.486 0.208 0.148 37.0 \n", "x -34.371 286.623 -142.353 209.821 20.340 14.404 290.0 \n", "\n", " ess_tail r_hat \n", "v 26.0 1.1 \n", "x 232.0 1.1 " ] }, "execution_count": 51, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_5, hdi_prob=0.89)" ] }, { "cell_type": "markdown", "id": "077d555e", "metadata": {}, "source": [ "### R Code 13.27" ] }, { "cell_type": "code", "execution_count": 52, "id": "97ee6912", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 50% (4000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 75% (6000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 2e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.02 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 2e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.02 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 3e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.03 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 2e-06 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 0.02 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " parameters {\n", " real v;\n", " real z;\n", " }\n", " \n", " model {\n", " v ~ normal(0, 3);\n", " z ~ normal(0, 1);\n", " }\n", " \n", " generated quantities {\n", " real x;\n", " \n", " x = z*exp(v);\n", " }\n", "\"\"\"\n", "\n", "posteriori = stan.build(model)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 53, "id": "3a716084", "metadata": {}, "outputs": [], "source": [ "model_13_6 = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", ")" ] }, { "cell_type": "code", "execution_count": 54, "id": "88f77684", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meansdhdi_5.5%hdi_94.5%mcse_meanmcse_sdess_bulkess_tailr_hat
v0.0552.991-4.4015.0180.0500.0483573.02181.01.0
z0.0180.998-1.5151.6630.0180.0153015.02616.01.0
x4.0283223.155-22.22030.35650.98236.0522943.02668.01.0
\n", "
" ], "text/plain": [ " mean sd hdi_5.5% hdi_94.5% mcse_mean mcse_sd ess_bulk \\\n", "v 0.055 2.991 -4.401 5.018 0.050 0.048 3573.0 \n", "z 0.018 0.998 -1.515 1.663 0.018 0.015 3015.0 \n", "x 4.028 3223.155 -22.220 30.356 50.982 36.052 2943.0 \n", "\n", " ess_tail r_hat \n", "v 2181.0 1.0 \n", "z 2616.0 1.0 \n", "x 2668.0 1.0 " ] }, "execution_count": 54, "metadata": {}, "output_type": "execute_result" } ], "source": [ "az.summary(model_13_6, hdi_prob=0.89)" ] }, { "cell_type": "markdown", "id": "7b18d661", "metadata": {}, "source": [ "#### 13.4.2 Non-centered chimpanzees" ] }, { "cell_type": "markdown", "id": "c3e1f001", "metadata": {}, "source": [ "### R Code 13.28" ] }, { "cell_type": "code", "execution_count": 55, "id": "d719f814", "metadata": {}, "outputs": [], "source": [ "# Don't have apapt_delta in pystan3, until today." ] }, { "cell_type": "markdown", "id": "24a26935", "metadata": {}, "source": [ "### R Code 13.29" ] }, { "cell_type": "markdown", "id": "75592884", "metadata": {}, "source": [ "$$ L_i \\sim Binomial(1, p_i) $$\n", "\n", "$$ logit(p_i) = \\bar{\\alpha} + z_{ACTOR[i]} \\sigma_\\alpha + x_{BLOCK[i]}\\sigma_\\gamma + \\beta_{TREATMENT[i]} $$\n", "\n", "To Actor:\n", "$$ \\bar{\\alpha} \\sim Normal(0, 1.5) $$\n", "\n", "$$ z_j \\sim Normal(0, 1) $$\n", "\n", "$$ \\sigma_\\alpha \\sim Exponential(1) $$\n", "\n", "\n", "\n", "To Block:\n", "\n", "$$ x_j \\sim Normal(0, 1) $$\n", "\n", "$$ \\sigma_\\gamma \\sim Exponential(1) $$\n", "\n", "\n", "To Treatment:\n", "\n", "$$ \\beta_j \\sim Normal(0, 0.5) $$\n", "\n", "\n", "Where, each actor is defined by:\n", "\n", "$$ \\alpha_j = \\bar{\\alpha} + z_j\\sigma_\\alpha $$\n", "\n", "and, each block is defined by:\n", "\n", "$$ \\gamma_j = x_j\\sigma_\\gamma $$" ] }, { "cell_type": "code", "execution_count": 56, "id": "09f1699e", "metadata": { "tags": [ "remove_output" ] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\u001b[32mBuilding:\u001b[0m found in cache, done.\n", "\u001b[36mSampling:\u001b[0m 0%\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (1/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (2/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (3/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 0% (4/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 1% (103/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 3% (202/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 4% (301/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 5% (400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 6% (500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 8% (600/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 9% (700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 10% (800/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 11% (900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 12% (1000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 14% (1100/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 15% (1200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 16% (1300/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 18% (1400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 19% (1500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 20% (1600/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 21% (1700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 22% (1800/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 24% (1900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 25% (2000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 26% (2100/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 28% (2200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 29% (2300/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 30% (2400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 32% (2600/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 34% (2700/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 35% (2800/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 38% (3000/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 39% (3100/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 40% (3200/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 42% (3400/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 56% (4500/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 70% (5600/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 86% (6900/8000)\n", "\u001b[1A\u001b[0J\u001b[36mSampling:\u001b[0m 100% (8000/8000)\n", "\u001b[1A\u001b[0J\u001b[32mSampling:\u001b[0m 100% (8000/8000), done.\n", "\u001b[36mMessages received during sampling:\u001b[0m\n", " Gradient evaluation took 0.000171 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.71 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 0.000152 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.52 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 0.000158 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 1.58 seconds.\n", " Adjust your expectations accordingly!\n", " Gradient evaluation took 0.000234 seconds\n", " 1000 transitions using 10 leapfrog steps per transition would take 2.34 seconds.\n", " Adjust your expectations accordingly!\n" ] } ], "source": [ "model = \"\"\"\n", " data {\n", " int N;\n", " int qty_chimpanzees;\n", " int qty_blocks;\n", " int qty_treatments;\n", " \n", " array[N] int pulled_left;\n", " array[N] int actor;\n", " array[N] int block;\n", " array[N] int treatment;\n", " }\n", " \n", " parameters {\n", " // To treatments\n", " vector[qty_treatments] beta;\n", " \n", " // To actors\n", " real bar_alpha;\n", " vector[qty_chimpanzees] z;\n", " real sigma_alpha;\n", " \n", " // To block\n", " vector[qty_blocks] x;\n", " real sigma_gamma;\n", " }\n", " \n", " model {\n", " vector[N] p;\n", " \n", " // priors\n", " beta ~ normal(0, 0.5); // treatment\n", " z ~ normal(0, 1); // actor\n", " x ~ normal(0, 1); // block \n", " \n", " bar_alpha ~ normal(0, 1.5); // Intercept to alpha (actor)\n", " \n", " sigma_alpha ~ exponential(1);\n", " sigma_gamma ~ exponential(1);\n", " \n", " // Link\n", " for (i in 1:N){\n", " p[i] = bar_alpha + z[ actor[i] ]*sigma_alpha + x[ block[i] ]*sigma_gamma + beta[ treatment[i] ];\n", " p[i] = inv_logit(p[i]);\n", " }\n", " \n", " // Linkelihood\n", " pulled_left ~ binomial(1, p);\n", " }\n", " \n", " generated quantities {\n", " vector[qty_chimpanzees] alpha;\n", " vector[qty_blocks] gamma;\n", " \n", " alpha = bar_alpha + z*sigma_alpha;\n", " gamma = x*sigma_gamma;\n", " }\n", "\n", "\"\"\"\n", "\n", "dat_list = df[['pulled_left', 'actor', 'block', 'treatment']].to_dict('list')\n", "dat_list['N'] = len(df)\n", "dat_list['qty_chimpanzees'] = len(df['actor'].unique())\n", "dat_list['qty_blocks'] = len(df['block'].unique())\n", "dat_list['qty_treatments'] = len(df['treatment'].unique())\n", "\n", "posteriori = stan.build(model, data=dat_list)\n", "samples = posteriori.sample(num_chains=4, num_samples=1000)" ] }, { "cell_type": "code", "execution_count": 57, "id": "3f1be292", "metadata": {}, "outputs": [], "source": [ "model_13_4_nc = az.from_pystan(\n", " posterior=samples,\n", " posterior_model=posteriori,\n", " observed_data=dat_list.keys()\n", ")" ] }, { "cell_type": "code", "execution_count": 58, "id": "adb86a60", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "az.plot_forest(model_13_4_nc, hdi_prob=0.89, combined=True, figsize=(17, 13))\n", "\n", "plt.axvline(x=0, ls='--', color='red')\n", "plt.grid(axis='y', color='white', ls='--', alpha=0.5)\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "755cc932", "metadata": {}, "source": [ "### R Code 13.30" ] }, { "cell_type": "code", "execution_count": 59, "id": "cfebc789", "metadata": {}, "outputs": [], "source": [ "# Extract features from ess\n", "\n", "# non-centered\n", "ess_nc = np.array(az.ess(model_13_4_nc, var_names=['alpha']).alpha.values)\n", "ess_nc = np.append(ess_nc, az.ess(model_13_4_nc, var_names=['beta']).beta.values)\n", "ess_nc = np.append(ess_nc, az.ess(model_13_4_nc, var_names=['gamma']).gamma.values)\n", "ess_nc = np.append(ess_nc, az.ess(model_13_4_nc, var_names=['bar_alpha']).bar_alpha.values)\n", "ess_nc = np.append(ess_nc, az.ess(model_13_4_nc, var_names=['sigma_alpha']).sigma_alpha.values)\n", "ess_nc = np.append(ess_nc, az.ess(model_13_4_nc, var_names=['sigma_gamma']).sigma_gamma.values)\n", "\n", "# centered\n", "ess_c = np.array(az.ess(model_13_4, var_names=['alpha']).alpha.values)\n", "ess_c = np.append(ess_c, az.ess(model_13_4, var_names=['beta']).beta.values)\n", "ess_c = np.append(ess_c, az.ess(model_13_4, var_names=['gamma']).gamma.values)\n", "ess_c = np.append(ess_c, az.ess(model_13_4, var_names=['bar_alpha']).bar_alpha.values)\n", "ess_c = np.append(ess_c, az.ess(model_13_4, var_names=['sigma_alpha']).sigma_alpha.values)\n", "ess_c = np.append(ess_c, az.ess(model_13_4, var_names=['sigma_gamma']).sigma_gamma.values)" ] }, { "cell_type": "code", "execution_count": 60, "id": "e4762b69", "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(17, 8))\n", "\n", "plt.scatter(ess_c, ess_nc)\n", "\n", "plt.plot([0, 2500], [0, 2500], ls='--', c='k', alpha=0.4)\n", "plt.text(650, 500, 'Identity line ($x=y$)')\n", "\n", "plt.title('Effective number samples')\n", "plt.xlabel('n_eff(centered)')\n", "plt.ylabel('n_eff(non-centered)')\n", "\n", "plt.show()" ] } ], "metadata": { "celltoolbar": "Tags", "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": "3.8.0" } }, "nbformat": 4, "nbformat_minor": 5 }