notebook/4.1_main-figure-experiments.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<h3>Runned on Corrado-Centai-Macbook.local</h3><h4>Last commit: b7bc819 Update experiments\n",
       "</h4><h4>Conda env: learnable-econ-abm</h4>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import display, HTML\n",
    "# from ipywidgets import interactive\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as ticker\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats\n",
    "import seaborn as sns\n",
    "import sklearn.metrics\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "from collections import namedtuple, defaultdict, Counter\n",
    "from glob import glob\n",
    "import itertools\n",
    "import os\n",
    "import pickle\n",
    "import re\n",
    "import socket\n",
    "import subprocess\n",
    "import sys\n",
    "\n",
    "plt.style.use('seaborn')\n",
    "# tf.compat.v1.logging.set_verbosity(tf.compat.v1.logging.ERROR)\n",
    "np.set_printoptions(suppress=True, precision=2, threshold=20)\n",
    "sys.path += ['../src/']\n",
    "\n",
    "display(HTML(\"<h3>Runned on %s</h3><h4>Last commit: %s</h4><h4>Conda env: %s</h4>\" % (\n",
    "    socket.gethostname(),\n",
    "    subprocess.run([\"git\",\"log\", \"-1\", \"--pretty=%h %s\"],\n",
    "        check=False, capture_output=True).stdout.decode() or 'none.',\n",
    "    (re.findall(\"\"\".*envs[\\\\\\/](.*?)[\\\\\\/]\"\"\", sys.executable) or [sys.executable])[0]\n",
    ")))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_experiments = pd.read_csv(\"../data/main-figure-experiments.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics = [\n",
    "    \"metrics.r2_score_M\",\n",
    "    \"metrics.r2_score_Nd\",\n",
    "    \"metrics.r2_score_P\",\n",
    "    \"metrics.r2_score_Db\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/corrado/.conda/envs/learnable-econ-abm/lib/python3.7/site-packages/ipykernel_launcher.py:5: FutureWarning: The default value of regex will change from True to False in a future version.\n",
      "  \"\"\"\n",
      "/Users/corrado/.conda/envs/learnable-econ-abm/lib/python3.7/site-packages/ipykernel_launcher.py:6: FutureWarning: The default value of regex will change from True to False in a future version.\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Model</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>original</td>\n",
       "      <td>$M$</td>\n",
       "      <td>-0.130794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>original</td>\n",
       "      <td>$M$</td>\n",
       "      <td>0.169643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>original</td>\n",
       "      <td>$M$</td>\n",
       "      <td>0.445487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>original</td>\n",
       "      <td>$M$</td>\n",
       "      <td>-0.005293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>original</td>\n",
       "      <td>$M$</td>\n",
       "      <td>0.222546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>learnable</td>\n",
       "      <td>$D^B$</td>\n",
       "      <td>0.760010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>learnable</td>\n",
       "      <td>$D^B$</td>\n",
       "      <td>0.754456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>learnable</td>\n",
       "      <td>$D^B$</td>\n",
       "      <td>0.782150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>learnable</td>\n",
       "      <td>$D^B$</td>\n",
       "      <td>0.779995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>learnable</td>\n",
       "      <td>$D^B$</td>\n",
       "      <td>0.782720</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>76 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Model variable     value\n",
       "0    original      $M$ -0.130794\n",
       "1    original      $M$  0.169643\n",
       "2    original      $M$  0.445487\n",
       "3    original      $M$ -0.005293\n",
       "4    original      $M$  0.222546\n",
       "..        ...      ...       ...\n",
       "71  learnable    $D^B$  0.760010\n",
       "72  learnable    $D^B$  0.754456\n",
       "73  learnable    $D^B$  0.782150\n",
       "74  learnable    $D^B$  0.779995\n",
       "75  learnable    $D^B$  0.782720\n",
       "\n",
       "[76 rows x 3 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_experiments[[\"Model\"] + metrics].melt(id_vars=['Model'], value_vars=metrics)\n",
    "df.variable = df.variable.astype('category')\n",
    "df.Model = df.Model.astype('category')\n",
    "df.variable = (df.variable\n",
    "               .str.replace(\"metrics.pearson_\", \"\")\n",
    "               .str.replace(\"metrics.r2_score_\", \"\")\n",
    "               .apply(lambda x: \"D^B\" if x == \"Db\" else \"D\" if x == \"Nd\" else x)\n",
    "               .apply(lambda x: '$' + x + '$')\n",
    "              )\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['$M$', '$D$', '$P$', '$D^B$'], dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.variable.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<svg  width=\"550\" height=\"55\"><rect x=\"0\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#4c72b0;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"55\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#dd8452;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"110\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#55a868;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"165\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#c44e52;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"220\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#8172b3;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"275\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#937860;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"330\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#da8bc3;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"385\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#8c8c8c;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"440\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#ccb974;stroke-width:2;stroke:rgb(255,255,255)\"/><rect x=\"495\" y=\"0\" width=\"55\" height=\"55\" style=\"fill:#64b5cd;stroke-width:2;stroke:rgb(255,255,255)\"/></svg>"
      ],
      "text/plain": [
       "[(0.2980392156862745, 0.4470588235294118, 0.6901960784313725),\n",
       " (0.8666666666666667, 0.5176470588235295, 0.3215686274509804),\n",
       " (0.3333333333333333, 0.6588235294117647, 0.40784313725490196),\n",
       " (0.7686274509803922, 0.3058823529411765, 0.3215686274509804),\n",
       " (0.5058823529411764, 0.4470588235294118, 0.7019607843137254),\n",
       " (0.5764705882352941, 0.47058823529411764, 0.3764705882352941),\n",
       " (0.8549019607843137, 0.5450980392156862, 0.7647058823529411),\n",
       " (0.5490196078431373, 0.5490196078431373, 0.5490196078431373),\n",
       " (0.8, 0.7254901960784313, 0.4549019607843137),\n",
       " (0.39215686274509803, 0.7098039215686275, 0.803921568627451)]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def define_palette():\n",
    "    (blue, orange, green, red, purple, \n",
    "     brown, pink, gray, gold, azure) = sns.color_palette('deep')\n",
    "    return {\n",
    "        '$M$': blue,\n",
    "        '$D^B$': red,\n",
    "        '$D$': green,\n",
    "        '$P$': brown,\n",
    "    }\n",
    "display(sns.color_palette('deep'))\n",
    "palette = define_palette()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir -p ../paper/fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x180 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x180 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "order = ['$M$', '$D^B$', '$D$', '$P$']\n",
    "for model in df.Model.unique():\n",
    "    sns.set_style(\"whitegrid\")\n",
    "    plt.figure(figsize=(7,2.5))\n",
    "    \n",
    "    ax = sns.boxplot(x=\"value\", y='variable',\n",
    "                     data=df[df.Model == model],\n",
    "                     whis=[0, 100],\n",
    "                     width=.6,\n",
    "                     order=order,\n",
    "                     palette=palette,\n",
    "                    )\n",
    "    sns.stripplot(x=\"value\", y='variable',\n",
    "                  data=df[df.Model == model],\n",
    "                  size=10, jitter=0.06,\n",
    "                  edgecolor=\"black\",\n",
    "                  linewidth=0.5,\n",
    "                  order=order,\n",
    "                  palette=palette,\n",
    "                 )\n",
    "\n",
    "    # metric_label = r'$\\rho(%s, \\hat{%s})$' % (label, label)\n",
    "#     plt.xlabel(\"Correlation with ground truth\")\n",
    "    plt.xlabel(\"Determination of ground truth ($R^2$)\")\n",
    "    plt.xlim((-.2, 1.05))\n",
    "    plt.ylabel(\"Observable        Latent\")\n",
    "    ax.yaxis.set_label_coords(-0.07, 0.47)\n",
    "    plt.title(model.title() + \" ABM as ground truth\")\n",
    "    plt.axhline(y=1.5, color='k', linestyle='-', alpha=0.1)\n",
    "\n",
    "    for ext in ('pdf', 'png'):\n",
    "        plt.savefig(f\"../paper/fig/estimate-{model}.{ext}\",\n",
    "                        bbox_inches='tight', dpi=200)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "learnable-econ-abm",
   "language": "python",
   "name": "learnable-econ-abm"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}