{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "49039775",
   "metadata": {},
   "source": [
    "# Inverse Probability of Treatment Weighting (IPTW)\n",
    "\n",
    "`IPTWWeighter` is a propensity-score-based weighter: rather than selecting a subset of the pool\n",
    "(as the genetic/LP/propensity matchers elsewhere in `pybalance` do), it **reweights every pool\n",
    "patient**. It:\n",
    "\n",
    "1. fits a propensity model $p(X) = P(\\mathrm{target} \\mid X)$ that classifies pool vs. target\n",
    "   patients on their covariates (`sklearn.linear_model.LogisticRegression` by default), then\n",
    "2. weights each pool patient by the odds $p / (1 - p)$.\n",
    "\n",
    "The target population keeps weight 1, so this is the **ATT estimand**, with the target playing the\n",
    "role of the fixed/reference (\"treated\") group -- the natural fit for this package's typical use\n",
    "case of reweighting a real-world data (RWD) pool to represent a trial population.\n",
    "\n",
    "IPTW:\n",
    "- **requires a patient-level target** (there's no pool-vs-target classification problem to fit\n",
    "  against a published, aggregate-only Table 1);\n",
    "- **does not guarantee exact balance** on any particular moment -- only asymptotically, and only\n",
    "  if the propensity model is well specified;\n",
    "- is nonetheless the most widely used and reported method for this kind of reweighting in the\n",
    "  observational literature (Rosenbaum & Rubin, 1983; Austin, 2011).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "82f52194",
   "metadata": {},
   "outputs": [],
   "source": [
    "import logging\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from pybalance.sim import generate_toy_dataset\n",
    "from pybalance.utils import MatchingData, MatchingHeaders\n",
    "from pybalance.visualization import plot_categoric_features, plot_numeric_features\n",
    "from pybalance.weighting import (\n",
    "    IPTWWeighter,\n",
    "    plot_iptw_propensity_distributions,\n",
    "    weighted_balance_table,\n",
    ")\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "# show effective-sample-size / trimming diagnostics logged during fit()\n",
    "logging.basicConfig(level=logging.INFO, format=\"%(message)s\", force=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39bb7786",
   "metadata": {},
   "source": [
    "## Synthetic data: two genuinely different, patient-level populations\n",
    "\n",
    "IPTW needs patient-level covariates on *both* sides so it can fit a pool-vs-target classifier. We\n",
    "use `generate_toy_dataset`'s pool and target populations directly, with no aggregation step."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a2f37129",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>pool</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>population size</th>\n",
       "      <th>N</th>\n",
       "      <td>2000.0000</td>\n",
       "      <td>300.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">gender</th>\n",
       "      <th>0.0</th>\n",
       "      <td>0.4830</td>\n",
       "      <td>0.553333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>0.5170</td>\n",
       "      <td>0.446667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">haircolor</th>\n",
       "      <th>0.0</th>\n",
       "      <td>0.3935</td>\n",
       "      <td>0.176667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>0.3045</td>\n",
       "      <td>0.403333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.3020</td>\n",
       "      <td>0.420000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"7\" valign=\"top\">age</th>\n",
       "      <th>mean</th>\n",
       "      <td>55.4200</td>\n",
       "      <td>46.710000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>13.1500</td>\n",
       "      <td>13.720000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>18.2700</td>\n",
       "      <td>18.720000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q25</th>\n",
       "      <td>46.6200</td>\n",
       "      <td>36.290000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median</th>\n",
       "      <td>57.4700</td>\n",
       "      <td>47.410000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q75</th>\n",
       "      <td>66.1700</td>\n",
       "      <td>55.930000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>74.9900</td>\n",
       "      <td>74.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"7\" valign=\"top\">weight</th>\n",
       "      <th>mean</th>\n",
       "      <td>88.7000</td>\n",
       "      <td>82.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>16.4400</td>\n",
       "      <td>19.110000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>50.2200</td>\n",
       "      <td>50.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q25</th>\n",
       "      <td>76.7700</td>\n",
       "      <td>65.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median</th>\n",
       "      <td>89.4800</td>\n",
       "      <td>81.130000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q75</th>\n",
       "      <td>101.1000</td>\n",
       "      <td>96.190000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>119.9500</td>\n",
       "      <td>119.660000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"7\" valign=\"top\">height</th>\n",
       "      <th>mean</th>\n",
       "      <td>159.0500</td>\n",
       "      <td>155.180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>19.5900</td>\n",
       "      <td>17.190000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>125.0400</td>\n",
       "      <td>125.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q25</th>\n",
       "      <td>142.1000</td>\n",
       "      <td>140.920000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median</th>\n",
       "      <td>158.7400</td>\n",
       "      <td>154.890000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>q75</th>\n",
       "      <td>175.3700</td>\n",
       "      <td>167.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>194.9500</td>\n",
       "      <td>194.230000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             pool      target\n",
       "population size N       2000.0000  300.000000\n",
       "gender          0.0        0.4830    0.553333\n",
       "                1.0        0.5170    0.446667\n",
       "haircolor       0.0        0.3935    0.176667\n",
       "                1.0        0.3045    0.403333\n",
       "                2          0.3020    0.420000\n",
       "age             mean      55.4200   46.710000\n",
       "                std       13.1500   13.720000\n",
       "                min       18.2700   18.720000\n",
       "                q25       46.6200   36.290000\n",
       "                median    57.4700   47.410000\n",
       "                q75       66.1700   55.930000\n",
       "                max       74.9900   74.600000\n",
       "weight          mean      88.7000   82.100000\n",
       "                std       16.4400   19.110000\n",
       "                min       50.2200   50.160000\n",
       "                q25       76.7700   65.520000\n",
       "                median    89.4800   81.130000\n",
       "                q75      101.1000   96.190000\n",
       "                max      119.9500  119.660000\n",
       "height          mean     159.0500  155.180000\n",
       "                std       19.5900   17.190000\n",
       "                min      125.0400  125.160000\n",
       "                q25      142.1000  140.920000\n",
       "                median   158.7400  154.890000\n",
       "                q75      175.3700  167.800000\n",
       "                max      194.9500  194.230000"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m = generate_toy_dataset(n_pool=2000, n_target=300, seed=11)\n",
    "pool_df = m.get_population(\"pool\").drop(columns=[m.population_col])\n",
    "target_df = m.get_population(\"target\").drop(columns=[m.population_col])\n",
    "\n",
    "headers = MatchingHeaders(numeric=[\"age\", \"weight\", \"height\"], categoric=[\"gender\", \"haircolor\"])\n",
    "matching_data = MatchingData(pool=pool_df, target=target_df, headers=headers)\n",
    "matching_data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd015b59",
   "metadata": {},
   "source": [
    "## Fit `IPTWWeighter`\n",
    "\n",
    "As with the other weighters, `match()` fits (if needed) and returns a `MatchingData` with every\n",
    "pool patient retained, plus a new `sample_weight` column. `weighted_balance_table()` compares\n",
    "each feature's target moment to the unweighted and weighted pool moment -- but, unlike an\n",
    "exact-balance weighter, don't expect the \"weighted\" column to land exactly on target."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2d0b37a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IPTWWeighter fit LogisticRegression. Effective sample size: 946.5 / 2000 pool patients.\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>feature</th>\n",
       "      <th>moment</th>\n",
       "      <th>target</th>\n",
       "      <th>unweighted_pool</th>\n",
       "      <th>weighted_pool</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>gender_1.0</td>\n",
       "      <td>mean</td>\n",
       "      <td>0.446667</td>\n",
       "      <td>0.517000</td>\n",
       "      <td>0.498286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>haircolor_1</td>\n",
       "      <td>mean</td>\n",
       "      <td>0.403333</td>\n",
       "      <td>0.304500</td>\n",
       "      <td>0.403873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>haircolor_2</td>\n",
       "      <td>mean</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.302000</td>\n",
       "      <td>0.414048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>age</td>\n",
       "      <td>mean</td>\n",
       "      <td>46.705231</td>\n",
       "      <td>55.417328</td>\n",
       "      <td>47.326641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>weight</td>\n",
       "      <td>mean</td>\n",
       "      <td>82.095245</td>\n",
       "      <td>88.704041</td>\n",
       "      <td>82.229043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>height</td>\n",
       "      <td>mean</td>\n",
       "      <td>155.177505</td>\n",
       "      <td>159.050598</td>\n",
       "      <td>153.852687</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       feature moment      target  unweighted_pool  weighted_pool\n",
       "0   gender_1.0   mean    0.446667         0.517000       0.498286\n",
       "1  haircolor_1   mean    0.403333         0.304500       0.403873\n",
       "2  haircolor_2   mean    0.420000         0.302000       0.414048\n",
       "3          age   mean   46.705231        55.417328      47.326641\n",
       "4       weight   mean   82.095245        88.704041      82.229043\n",
       "5       height   mean  155.177505       159.050598     153.852687"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weighter = IPTWWeighter(matching_data, verbose=True)\n",
    "matched = weighter.match()\n",
    "\n",
    "weighted_balance_table(weighter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "936d3cfc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Effective sample size: 946.5 / 2000 pool patients\n",
      "Propensity model: LogisticRegression(max_iter=1000)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    2000.000000\n",
       "mean        0.146556\n",
       "std         0.154660\n",
       "min         0.006937\n",
       "25%         0.052532\n",
       "50%         0.094784\n",
       "75%         0.184358\n",
       "max         1.642370\n",
       "Name: sample_weight, dtype: float64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(f\"Effective sample size: {weighter.effective_sample_size():.1f} / {len(pool_df)} pool patients\")\n",
    "print(f\"Propensity model: {weighter.propensity_model}\")\n",
    "matched.get_population(matched.pool_name)[\"sample_weight\"].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fd1dc9e",
   "metadata": {},
   "source": [
    "## Checking overlap: propensity score and weight distributions\n",
    "\n",
    "IPTW relies on **positivity**: every pool patient needs a realistic chance of looking like the\n",
    "target on the fitted model. A propensity score distribution piled up near 0 or 1, or a handful of\n",
    "huge weights, is the practical warning sign -- the reweighting is leaning hard on a small,\n",
    "possibly unrepresentative slice of the pool (the same concern `effective_sample_size()` summarizes\n",
    "in one number).\n",
    "\n",
    "`plot_iptw_propensity_distributions()` is the weighting counterpart to\n",
    "`pybalance.propensity.plot_propensity_score_match_distributions()`: instead of a matched subset's\n",
    "\"before\" vs. \"after\" panels, it shows the same pool/target propensity scores counted equally\n",
    "(\"before\") vs. counted by the fitted IPTW weight (\"after\") -- the weighted pool histogram should\n",
    "move towards the target's."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5c03e336",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_iptw_propensity_distributions(weighter);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac79849c",
   "metadata": {},
   "source": [
    "Both `plot_numeric_features` and `plot_categoric_features` accept a `weights=` column, so the\n",
    "unweighted pool (every patient counted equally) and the IPTW-weighted pool can be plotted\n",
    "side by side against the same target, to see what the reweighting actually bought."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f06686ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 550x350 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(5.5, 3.5))\n",
    "ax.hist(weighter.weights, bins=30)\n",
    "ax.set_title(r\"Pool weights, $\\hat{p} / (1 - \\hat{p})$\")\n",
    "ax.set_xlabel(\"weight\")\n",
    "fig.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4529cba7",
   "metadata": {},
   "source": [
    "To see the effect on the covariates themselves (rather than the propensity score), put the\n",
    "unweighted pool, the IPTW-weighted pool and the target into a single `MatchingData` -- the same\n",
    "trick the matching demos use to show \"pool (prematch)\" alongside \"pool\" and \"target\" in one plot --\n",
    "and pass `weights=` through to `plot_numeric_features`/`plot_categoric_features`. The unweighted\n",
    "pool copy gets weight 1, so a single call draws all three distributions together instead of two\n",
    "separate before/after figures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d7fc0250",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x450 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "before_pool = matching_data.get_population(matching_data.pool_name).copy()\n",
    "before_pool[weighter.weight_col] = 1.0\n",
    "\n",
    "combined = matched.copy()\n",
    "combined.append(before_pool, name=\"pool (before)\")\n",
    "hue_order = [\"pool (before)\", matching_data.pool_name, matching_data.target_name]\n",
    "\n",
    "plot_numeric_features(\n",
    "    combined, weights=weighter.weight_col, hue_order=hue_order, include_only=headers.numeric, col_wrap=3\n",
    ")\n",
    "plt.suptitle(\"Pool vs. target, before vs. after IPTW weighting (numeric)\", y=1.05);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1d8f9dea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x450 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_categoric_features(\n",
    "    combined, weights=weighter.weight_col, hue_order=hue_order, include_only=headers.categoric, col_wrap=2\n",
    ")\n",
    "plt.suptitle(\"Pool vs. target, before vs. after IPTW weighting (categoric)\", y=1.05);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd1e775e",
   "metadata": {},
   "source": [
    "## Trimming extreme weights\n",
    "\n",
    "The main practical failure mode of IPTW is a small number of extreme weights, driven by pool\n",
    "patients whose covariates make the model almost certain they're pool (weight -> 0) or almost\n",
    "certain they're target (weight -> large). `trim_quantiles=(low, high)` clips fitted weights at the\n",
    "given quantiles -- a standard bias/variance trade-off: a little exact balance is sacrificed in\n",
    "exchange for a much more stable, lower-variance set of weights."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "62900099",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IPTWWeighter fit LogisticRegression. Effective sample size: 1148.0 / 2000 pool patients. Trimmed 200 extreme weights.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Untrimmed: ESS=946.5, max weight=1.64\n",
      "Trimmed:   ESS=1148.0, max weight=0.46, n_trimmed=200\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgkAAAFMCAYAAABBHy+6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8ekN5oAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAxXElEQVR4nO3dB5hTZb7H8T+9d6lLUekoZQEFBBRhZETWhQt7RURkd1EUAQWWIooIWIZFBZRLWRUpd0EEV1BREUREqVJkRUCkKSjNVelLz31+792TJxkOTCEzk8x8P8+TJ5PkJDlvJjnnd95y3myBQCBgAAAAiWRPfAcAAIAQEgAAgC9CAgAA8EVIAAAAvggJAADAFyEBAAD4IiQAAABfOS0GXbhwwfbt22eFChWybNmyZfTqAAAQM3R6pGPHjlm5cuUse/bsmS8kKCBUqFAho1cDAICYtXfvXitfvnzmCwmqQfAKWLhw4YxeHQAAYsbRo0fdgba3L810IcFrYlBAICQAAJByyWmup+MiAADwRUgAAAC+CAkAACDz9EkAEP3Onz9vZ8+ezejVALKcXLlyWY4cOSLyWoQEABEfg33gwAE7fPhwRq8KkGUVLVrUypQpc8XnEiIkAIgoLyCUKlXK8ufPzwnPgHQO6SdPnrRDhw6522XLlr2i1yMkAIhoE4MXEEqUKJHRqwNkSfny5XPXCgr6LV5J0wMdFwFEjNcHQTUIADKO9xu80n5BhAQAEUcTA5A5foOEBAAA4Is+CSGGvL0pVc9L6FA74usCAEBGIyQAiOoQnlrRHt5VHTxv3jxr3769xYI//vGPrlPq/PnzM3pVkI5obgCACPj000/djj+554fYv3+/tWnTJs3XC0jXkPDjjz/avffe64Y3aZhF7dq1bd26dWFjNIcNG+bGZurxuLg42759e9hr/PLLL9alSxc3g6NO+NC9e3c7fvz4FRUEAGLBmTNn3LVOdJMnT56MXh0gciHh119/taZNm7pTPn744Ye2ZcsWe/HFF61YsWLBZUaPHm0vv/yyTZ482dasWWMFChSw+Ph4O3XqVHAZBYTNmzfb4sWLbcGCBfbZZ59Zjx49UrIqABBRV199tY0bNy7svnr16tnw4cPd36oleO211+y//uu/3PCyqlWr2rvvvuse++677+zWW291f2t7qGVVPS8tWrSw3r17W9++fe2qq65y20Pv9byqez1ft+fMmWPNmzd3B1g33HCDffvtt7Z27Vpr2LChFSxY0NU8/PTTT8H103uoueK5556z0qVLu4OukSNH2rlz52zgwIFWvHhxK1++vE2dOjWsXHv37rW77rrLLa9l2rVr59Yh9HwX/fv3d4/rgHDQoEHuABBZT4pCwl//+lerUKGC+8LdeOONds0111jr1q2tcuXK7nF9ifQjGzp0qPvS1alTx2bMmGH79u0L/hi2bt1qCxcudD+2Ro0aWbNmzWz8+PE2e/ZstxwARKsRI0a4netXX31ld9xxhzvgUc2otov/+Mc/3DLbtm1zTQkvvfRS8HnTp0+33Llz24oVK9wB1KU89dRTbvu5YcMGy5kzp91zzz1uB63X+vzzz23Hjh2upjbUJ5984radOtgaM2aMe43f/e53LqzoQO2hhx6yBx980H744YfguHkFlUKFCrnX1DopgNx+++3BWg4d/E2bNs1ef/11W758uSuj+k8g60lRSFBqVqL97//+b3cWp9/+9rf26quvBh/fvXu3OyWrmhg8RYoUcWFg1apV7raulU71Oh4tnz17dveF9nP69Gk7evRo2AUA0puO3Dt37mxVqlRxR+9qJv3iiy/cGe10RC7aNqopQds+j2odVMtavXp1d7mUAQMGuB14zZo17dFHH7X169fbk08+6Wpwtb1V0+zSpUvDnqP3Ve2tXvfPf/6zu9ZpeR9//HH3vkOGDHEBRTt7efPNN+3ChQvuQE3NxXovHfjt2bPH9asQHezpeR06dHCPK9iElgdZR4pCwq5du2zSpEnui/fRRx9Zz5497ZFHHnEpWRQQRNVeoXTbe0zX+hGFUmLWF91bJrGEhAT3BfUuSu0AkN5UO+pRU6r6VXnnyL+cBg0apPj1ve2oduSh9yV+v+uuu84dZIUuE/ocBRg1GXjP++c//+lqJFSToBoEXbT9VZPwzp077ciRI64mRAd3odvo0AM7ZB0pGgKp9KkvihK0KNl+/fXXLmV269YtrdbRJVq1j3lUk0BQABBJ2tEmbndPfEpb9ccKpX4E2i4mRYEiOUJf3ztjXuL7Er+f3zpdbj1V+6HQMnPmzIvev2TJkslaT2QdKapJ0IiFWrVqhd2nqihVU4mq2OTgwYNhy+i295iuEydhdbJRm5e3TGLqAazEHnoBgEjSDlJH0KEHI2pCTS5V6Xud/qJZ/fr13Ygz1eiq2ST04tXWalsf2vyrbbSaPpD1pCgkqF1MnXJCqfdtpUqV3N/qyKgd/ZIlS8J+aPqyNWnSxN3WtcYRh37h1PFGKTe0egsA0lPLli3tf//3f11nvk2bNrna0ZTMnqftoI7YNWJLIxCidVi3OltqlIU6l6usCkLqi6CmY69zo/pDjBo1ynU4/+abb+zhhx9O9vkfkIWbG/r162c33XSTa25QD1912HnllVfcRfQD0TCfZ555xvVbUGhQp5ty5coFzyqmmgf1on3ggQdcM4Wq8zQ86O6773bLAcicov0MiGrW1A5TIwN0NP3000+nqCbhN7/5jRv98Nhjj9mf/vQnu++++9wIgWij4ZsaCTF48GDXMfHYsWNu3Vu1ahWspf3LX/7ialUUlNQMow6RGvqp/grIWrIFUjj4VSlZPyZVVykEqK+AdvgevZyG4Cg4KHlqiOPEiROtWrVqwWXUtKBg8N5777kvYMeOHV3vXHWgSQ7VTuhHrC9sJJsemLsBuDLq/KYdq7YNefPmzejVAbKsU5f5LaZkH5rikBANCAlAdCIkAJkrJDB3AwAA8EVIAAAAvggJAADAFyEBAAD4IiQAAABfhAQAAOCLkAAAAHwREgAghYYPH2716tWzWPHdd9+5M+Ju3LjxssvptPs6tb7Owpge75fe03x7Z/6Ntu/PH1O4blu2bLHy5cvbiRMnLKpOywwAqfbeo+n7fne+lKLFW7Ro4Tbc48aNS3LZAQMGWJ8+fSyz0dl0VS5NI+3t7HUynsRWrVpljRs3Du7gdHZdzfOQGsuWLbN7773X9u7d68KJTuU/b948NxGgZhp+6aWX7IYbbggur/ebPn162GvEx8fbwoULw9b5yy+/jIogly1bNlee0BBwpd8fTbSoz3/MmDHu80pL1CQAQDLpBLWaEVGnkC9RooRlJprNV6fd1044sY8//tjN5eBdNNV0pLzzzjt25513ur/vv/9+W7x4sZtoS5NstW7d2uLi4uzHH38Me47m/wldnzfeeMNiScEIfH80P8ikSZPc9zEtERIAZHnaMeqIVketOvLTRUekmh1Rf3/44Ydux6hp65cvX37J6mJNfle6dGkrWrSojRw50m3ABw4caMWLF3fVw1OnTr2oSn7OnDnWvHlzy5cvnzti1sy6a9eutYYNG7qdSZs2bdyskqFee+01N1meTrdbo0YNNz9OKE2+p6NwPa7X0VF1UrQedevWdZM9JaYdmpohvEuuXLnc/focdFSvHb33uekz8+zatctuvfVWN6mUXls1EIm9++679vvf/97+/e9/2z/+8Q8bPXq03XzzzW7qar2+rrUzDKX/Q+j6FCtWLPiYV/Oh8mt9VEMU6oUXXnBTYatMvXr1cpMMXsrw//yf//a3v1mFChVcOTS5YeKJrl5//XW77rrr3HrptTU3kVx99dXuWpNjaV2820k1V2lW5ISEBFcWfS/02b311lthy9x2221uHiR9b9MSIQFAlqdwoGnsNVmdd3SqnYJHMztq6uStW7danTp1fF9DU97v27fPzbCoamBNdKcZJbUDW7NmjT300EP24IMPBqdj9mi5oUOH2oYNGyxnzpx2zz332KBBg9w6aSrnHTt22LBhw4LLz5w5091+9tln3foomKjK2auC1xTVel9VSa9fv97tkFS9nRS9lwKFH+3ES5Uq5Sbs007do9fVTjP0yF4zBXueeOIJt4z6JmiSv86dO4cd+W7evNk1K2iabt1//vz5i+YZ0E5SwSyUgojWp3r16tazZ0/7+eefwwJSaO3H22+/HXxs6dKltnPnTnetz0uzdCY1U+eOHTtcgNKEhGrSUODS1NkeBRiFjR49erjaD30+CjaisCcKh1oX73ZSFBBmzJjhZkrWZ6QZmNUkExoIcufO7YKG/m9piT4JALI8TXajja6OFHVkmphqBXTkdjmqLdBstprZVjsvHRGfPHnSHn/88WB7v4KGdnh333138HnaiapNXR599FG3I12yZIk1bdrU3de9e/ewHZlCxYsvvuimeRYdbaojm452NbXzrFmz3JHolClT3A5XR7gKJtqZXs73339/UUhQTYbeS+uiculIXzUm6n+g4KDHtRM/ffq07+emsrVt29b9rWm0tS7a6ar2Q1QDobLrs9dFQU1TdKuWRDUyakZQ7YO30xUFEpVd5dYOX5+valu0XI4cOaxkyZJhtR+hFNj+53/+xy2nddC66bMOncnYb6KkGTNmBGtYxo8f756nz0Wv/8wzz7iptfW/83h9KLx1Uc2S3+fjR5+lgp9Cjj4Pufbaa933Rv/jW265JbhsuXLl3P8tLRESIoDZI4HM7VJH2KG0A9SO1KOd3PXXXx+8rR2Tdlw6cg4VWjOh50jt2rXD7vOeo97s2jEqOITu2HQUrqAjXm1H6BG5t7O5HFX3Jz6Kv+qqq6x///5hOz/Vljz//PMuJCQltGyqhheVJTQkeFXzor4If/7zn90OWZ9X/fr1XWhSjYgnNGDpc9J7VK5c2dUutGrVKsn/kV43dJ109H85FStWDGuC0WepEKaRIPp/6/NI6n1TQiFK4TJxKD1z5oxrQgmlgKZl0xIhAQCSUKBAgSSX8drpPWqD9rtPO5hLPU+P+93nPUdNCfLqq69ao0aNwl4ndOeXGgoEv/76a5LL6X3VuTA5/MrmlUXV76q692oaRDt7VakrDGk6Y+3EO3Xq5I6kL0WPad21c01qZ52c/0dK5MuXzyLN+x+///77F/UPUZ+HUOqToM8sLdEnAQD+08arNvFoploFVTGrQ6Cq4EMvXoc9VdV/9dVXrprcs3r16iRfW0eparZIivoXeLUCV/K5qY1f/RfUTOMXyvQeCi0fffSRtWvX7pKvo6YU9Unw1knrI5H6X+7Zs8fVFoR+ll6TkoaKqjOimiwuF0xSsi7qS6IwoPdN/D8O7ScjX3/99UW1C5FGTQIA/KcnujoYatSB2tr9dl7RQG37jzzyiGteUPu82rDXrVvndqhqGlDHR3UYVHOE+kGoPOrRnxT1DdAQRO3QvFoJde7TTtfbEakToHrya3RF6OemHbmq39Wc4jV7JMUb1RBKr6NhptoBq2ZAI0PUNKHhft5RtsrfsWNH18avphd18tQO1OvXoQ6NOsJXJ0ONKFETSnLXyU/evHldXw99hqrd0GevzppeHwN1DFWnVL2v+kboXA8rVqwIngfBCxHq16Gdf+hIDD8KHurLoc6KquVQZ1GNptBrFi5c2K2L6P+qoaEaIpqWqEkAgP90stPOUUdy6nCmI7lopB25dtLqMa82eXVkU8dGryZBAUdH6Wpr185dgeGvf/1rkq+rHZxGV6jDXCh1JNTwTzUzqA/Bm2++Gdxpi8KIdurqt6HPTTuzpKg5QTvOxCFBO0ONFFAwuO+++9wOUsHBaybQ/0e1JHqeRkuob4bWTT38vap4lUEdSNXJT7Uul6uFSI4qVaq4jpJ33HGHO2+D+kCEDjnVTlsn4NJ96vOgkSXbt28PPq4OjmqeUS1Aco/69ZlrxIpGOahmSGFQzQ+hJ7ZSp06tT6VKlSwtZQsotsUYpTklQ32hlKwyugNiatFxEZmNqrh3797tNmaJO8Eh+k2YMMEd4WvHnJZUI6Fhn8lp3shIw4cPdyM5oun00l4nxqpVq7qRLN4omJT8FlOyD6W5AQDg6DwOOsWyqsy9UzOnBdV2JKd2A/5Uy6Whn5cKCJFESAAABKvq1TyR1lRNjtTzOjKmB/okAABwieaGjVHW1JDeCAkAAMAXIQFAxMVgf2ggUwlE6DdISAAQMd5QtbQ+VSyAy/N+g4nPMplSdFwEEDEax67JbLy5BjRhknc6XgDpU4OggKDfoH6LV3q6bkICgIjyzkSXeCIjAOknJTNPXg4hAUBEqeZA59HXaWrPnj2b0asDZDm5cuW64hoEDyEBQJrQRipSGyoAGYOOiwAAwBchAQAAXHlI0Nmn1N4YetFsXaETSmgGL00XqnNzazrPgwcPXnTO6bZt27pez2qz1FSg586dS8lqAACAdJDiPgmaCjN0KlGd69uj+a81neXcuXPdDFO9e/d2U2x6U4dqnnIFBPW4XLlype3fv99NB6pOFs8991ykygQAADIiJCgU+A2r0JSTU6ZMcVNXtmzZ0t2n+c41F/bq1autcePGtmjRIjc1qEJG6dKlrV69em7e7MGDB7taity5c/u+5+nTp90ldJpLAAAQZX0Stm/fbuXKlbNrr73WunTp4poPZP369W64U1xcXHBZNUVUrFjRVq1a5W7runbt2i4geOLj491Of/PmzZd8z4SEBFcz4V0qVKiQ0tUGAABpGRIaNWpk06ZNs4ULF9qkSZNs9+7d1rx5czf3+IEDB1xNgE7gEEqBQI+JrkMDgve499ilDBkyxNVUeJe9e/emZLUBAEBaNze0adMm+HedOnVcaKhUqZLNmTPH8uXLZ2klT5487gIAAGJkCKRqDapVq2Y7duxw/RTOnDljhw8fDltGoxu8Pgy6TjzawbsdidNHAgCAKAkJx48ft507d7pTsDZo0MCNUliyZEnw8W3btrk+C02aNHG3db1p06awc7ovXrzYChcubLVq1bqSVQEAABnZ3DBgwAC78847XRPDvn377KmnnnKnXe3cubPrUNi9e3fr37+/FS9e3O34+/Tp44KBRjZI69atXRjo2rWrjR492vVDGDp0qDu3As0JAADEcEj44YcfXCD4+eefrWTJktasWTM3vFF/y9ixYy179uzuJEoasqiRCxMnTgw+X4FiwYIF1rNnTxceChQoYN26dbORI0dGvmQAAOCKZAto8ukYoyGTqrnQSAfVWETKkLc3WXpK6FA7Xd8PAICjKdiHMncDAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAA4IuQAAAAfBESAACAL0ICAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAA4IuQAAAAfBESAACAL0ICAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAA4IuQAAAAfBESAACAL0ICAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAAEPmQMGrUKMuWLZv17ds3eN+pU6esV69eVqJECStYsKB17NjRDh48GPa8PXv2WNu2bS1//vxWqlQpGzhwoJ07d+5KVgUAAERLSFi7dq397W9/szp16oTd369fP3vvvfds7ty5tmzZMtu3b5916NAh+Pj58+ddQDhz5oytXLnSpk+fbtOmTbNhw4ZdWUkAAEDGh4Tjx49bly5d7NVXX7VixYoF7z9y5IhNmTLFxowZYy1btrQGDRrY1KlTXRhYvXq1W2bRokW2ZcsW+/vf/2716tWzNm3a2NNPP20TJkxwwQEAAMRwSFBzgmoD4uLiwu5fv369nT17Nuz+GjVqWMWKFW3VqlXutq5r165tpUuXDi4THx9vR48etc2bN/u+3+nTp93joRcAAJC2cqb0CbNnz7YNGza45obEDhw4YLlz57aiRYuG3a9AoMe8ZUIDgve495ifhIQEGzFiREpXFQAApFdNwt69e+3RRx+1mTNnWt68eS29DBkyxDVleBetBwAAiKKQoOaEQ4cOWf369S1nzpzuos6JL7/8svtbNQLqV3D48OGw52l0Q5kyZdzfuk482sG77S2TWJ48eaxw4cJhFwAAEEUhoVWrVrZp0ybbuHFj8NKwYUPXidH7O1euXLZkyZLgc7Zt2+aGPDZp0sTd1rVeQ2HDs3jxYrfjr1WrViTLBgAA0qtPQqFChez6668Pu69AgQLunAje/d27d7f+/ftb8eLF3Y6/T58+Lhg0btzYPd66dWsXBrp27WqjR492/RCGDh3qOkOqxgAAAMRox8WkjB071rJnz+5OoqRRCRq5MHHixODjOXLksAULFljPnj1deFDI6Natm40cOTLSqwIAAK5AtkAgELAYoyGQRYoUcZ0YI9k/Ycjbmyw9JXSona7vBwDA0RTsQ5m7AQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAFceEiZNmmR16tSxwoULu0uTJk3sww8/DD5+6tQp69Wrl5UoUcIKFixoHTt2tIMHD4a9xp49e6xt27aWP39+K1WqlA0cONDOnTuXktUAAADRFhLKly9vo0aNsvXr19u6deusZcuW1q5dO9u8ebN7vF+/fvbee+/Z3LlzbdmyZbZv3z7r0KFD8Pnnz593AeHMmTO2cuVKmz59uk2bNs2GDRsW+ZIBAIArki0QCASu5AWKFy9uzz//vP3hD3+wkiVL2qxZs9zf8s0331jNmjVt1apV1rhxY1fr8Lvf/c6Fh9KlS7tlJk+ebIMHD7affvrJcufOnaz3PHr0qBUpUsSOHDniajQiZcjbmyw9JXSona7vBwDA0RTsQ1PdJ0G1ArNnz7YTJ064ZgfVLpw9e9bi4uKCy9SoUcMqVqzoQoLounbt2sGAIPHx8W6FvdoIP6dPn3bLhF4AAEDaypnSJ2zatMmFAvU/UL+DefPmWa1atWzjxo2uJqBo0aJhyysQHDhwwP2t69CA4D3uPXYpCQkJNmLECEtv7X8YHXZ7fvlB6b4OAABklBTXJFSvXt0FgjVr1ljPnj2tW7dutmXLFktLQ4YMcdUi3mXv3r1p+n4AACAVNQmqLahSpYr7u0GDBrZ27Vp76aWXrFOnTq5D4uHDh8NqEzS6oUyZMu5vXX/xxRdhr+eNfvCW8ZMnTx53AQAAMXSehAsXLrg+AwoMuXLlsiVLlgQf27ZtmxvyqOYJ0bWaKw4dOhRcZvHixa7jhJosAABAjNYkqNq/TZs2rjPisWPH3EiGTz/91D766CPXU7J79+7Wv39/N+JBO/4+ffq4YKCRDdK6dWsXBrp27WqjR492/RCGDh3qzq2QFWsKrmQ0BSMjAABRFRJUA3DffffZ/v37XSjQiZUUEG677Tb3+NixYy179uzuJEqqXdDIhYkTJwafnyNHDluwYIHry6DwUKBAAdenYeTIkZEvGQAASL+QMGXKlMs+njdvXpswYYK7XEqlSpXsgw8+SMnbAgCADMDcDQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwBchAQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAALjykJCQkGA33HCDFSpUyEqVKmXt27e3bdu2hS1z6tQp69Wrl5UoUcIKFixoHTt2tIMHD4Yts2fPHmvbtq3lz5/fvc7AgQPt3LlzKVkVAAAQTSFh2bJlLgCsXr3aFi9ebGfPnrXWrVvbiRMngsv069fP3nvvPZs7d65bft++fdahQ4fg4+fPn3cB4cyZM7Zy5UqbPn26TZs2zYYNGxbZkgEAgCuSMyULL1y4MOy2du6qCVi/fr3dfPPNduTIEZsyZYrNmjXLWrZs6ZaZOnWq1axZ0wWLxo0b26JFi2zLli328ccfW+nSpa1evXr29NNP2+DBg2348OGWO3fui9739OnT7uI5evRo6ksMAADSvk+CQoEUL17cXSssqHYhLi4uuEyNGjWsYsWKtmrVKndb17Vr13YBwRMfH+92/Js3b75kM0eRIkWClwoVKlzJagMAgLQMCRcuXLC+ffta06ZN7frrr3f3HThwwNUEFC1aNGxZBQI95i0TGhC8x73H/AwZMsQFEu+yd+/e1K42AABIi+aGUOqb8PXXX9vy5cstreXJk8ddAABAlNck9O7d2xYsWGBLly618uXLB+8vU6aM65B4+PDhsOU1ukGPecskHu3g3faWAQAAMRYSAoGACwjz5s2zTz75xK655pqwxxs0aGC5cuWyJUuWBO/TEEkNeWzSpIm7retNmzbZoUOHgstopEThwoWtVq1aV14iAACQ/s0NamLQyIV33nnHnSvB60OgzoT58uVz1927d7f+/fu7zoza8ffp08cFA41sEA2ZVBjo2rWrjR492r3G0KFD3WvTpAAAQIyGhEmTJrnrFi1ahN2vYY5//OMf3d9jx4617Nmzu5MoadiiRi5MnDgxuGyOHDlcU0XPnj1deChQoIB169bNRo4cGZkSAQCA9A8Jam5ISt68eW3ChAnucimVKlWyDz74wGJd+x9GB/+eX35Qhq4LAABRM7ohKwoNBQAAZHZM8AQAAHwREgAAgC+aG2LUkLc3pep5CR1qR3xdAACZEzUJAADAFyEBAAD4orkhnUZCMEQSABBrCAkRwvBIAEBmQ3MDAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAA4IuQAAAAfBESAACAL0ICAADwRUgAAAC+CAkAAMAXIQEAAPgiJAAAAF+EBAAA4IupojNoKun55Qdl2LoAAJAc1CQAAABfhAQAAOCLkAAAAHwREgAAgC9CAgAA8EVIAAAAvggJAAAgMiHhs88+szvvvNPKlStn2bJls/nz54c9HggEbNiwYVa2bFnLly+fxcXF2fbt28OW+eWXX6xLly5WuHBhK1q0qHXv3t2OHz+e0lUBLvbeo+EXAED6hYQTJ05Y3bp1bcKECb6Pjx492l5++WWbPHmyrVmzxgoUKGDx8fF26tSp4DIKCJs3b7bFixfbggULXPDo0aNH6ksBAAAy/oyLbdq0cRc/qkUYN26cDR061Nq1a+fumzFjhpUuXdrVONx99922detWW7hwoa1du9YaNmzolhk/frzdcccd9sILL7gaCgAAkMn6JOzevdsOHDjgmhg8RYoUsUaNGtmqVavcbV2ricELCKLls2fP7moe/Jw+fdqOHj0adgEAADE0d4MCgqjmIJRue4/pulSpUuErkTOnFS9ePLhMYgkJCTZixAjLrHM5pOc8DkPe3pSq5yV0qB3xdQEARLeYGN0wZMgQO3LkSPCyd+/ejF4lAAAyvYiGhDJlyrjrgwcPht2v295juj506FDY4+fOnXMjHrxlEsuTJ48bCRF6AQAAMRQSrrnmGrejX7JkSfA+9R9QX4MmTZq427o+fPiwrV+/PrjMJ598YhcuXHB9FwAAQIz2SdD5DHbs2BHWWXHjxo2uT0HFihWtb9++9swzz1jVqlVdaHjyySfdiIX27du75WvWrGm33367PfDAA26Y5NmzZ613795u5AMjGwAAiOGQsG7dOrv11luDt/v37++uu3XrZtOmTbNBgwa5cynovAeqMWjWrJkb8pg3b97gc2bOnOmCQatWrdyoho4dO7pzKwAAgBgOCS1atHDnQ7gUnYVx5MiR7nIpqnWYNWtWSt8aAACko5gY3QAAANIfIQEAAKT9yZRw5SdWSu+TKyUXJ2ECgKyHmgQAAOCLmoQoFAs1CwCAzI+aBAAA4IuQAAAAfNHcEANofgAAZARqEgAAgC9CAgAA8EVIAAAAvggJAADAFyEBAAD4IiQAAABfDIFEmmLOBwCIXYSETHYOhcQ4pwIAILUICZkwGFxuWUIDACC56JMAAAB8ERIAAIAvmhuQqazZ/UvY7fnJ7DhJR0kAuBg1CQAAwBc1CVlMaEdGOjECAC6HkIBMdX6F9hFfEwDIuggJWRjnWAAAXA4hAanC+RcAIPOj4yIAAPBFTQIigpoFAMh8CAkAE1EBgC9CAiIyR0RKXitxLQO1EAAQnQgJSPeAkdLwEc0hghoIAJkZIQHIAIQLALEgw0LChAkT7Pnnn7cDBw5Y3bp1bfz48XbjjTdm1OoAMYFwASDTh4Q333zT+vfvb5MnT7ZGjRrZuHHjLD4+3rZt22alSpXKiFVCFEmqOYJTSwNAJg4JY8aMsQceeMD+9Kc/udsKC++//769/vrr9thjj2XEKiFGRbJzJaKj1iO1UltbQu0MEEUh4cyZM7Z+/XobMmRI8L7s2bNbXFycrVq1yvc5p0+fdhfPkSNH3PXRo0cjum6nTx4Pu33i1JmIvj7SX+L/aVbV/+/+v63MJL3LGOntT1KGv7vZYsHw31+X0asQtYan8n8Y6c/U++4GAoHoCwn/+te/7Pz581a6dOmw+3X7m2++8X1OQkKCjRgx4qL7K1SoYGlpbJq+OtLHnIxeAWRSbB/88bnEzmd67NgxK1KkSOyPblCtg/oweC5cuGC//PKLlShRwrJly3bFiUphY+/evVa4cGGLdZQnumWm8mSmsgjliV6ZqSzRUB7VICgglCtXLsll0z0kXHXVVZYjRw47ePBg2P26XaZMGd/n5MmTx11CFS1aNKLrpX9UZvjyeShPdMtM5clMZRHKE70yU1kyujxJ1SBk2ARPuXPntgYNGtiSJUvCagZ0u0mTJum9OgAAIJqaG9R00K1bN2vYsKE7N4KGQJ44cSI42gEAAGTRkNCpUyf76aefbNiwYe5kSvXq1bOFCxde1JkxPagZ46mnnrqoOSNWUZ7olpnKk5nKIpQnemWmssRaebIFkjMGAgAAZDnp3icBAADEBkICAADwRUgAAAC+CAkAACDrhgRNS3311Vdb3rx53ayTX3zxxWWXnzt3rtWoUcMtX7t2bfvggw8sVsvz6quvWvPmza1YsWLuojkykip/tP9/PLNnz3Zn3Gzfvr3FalkOHz5svXr1srJly7qeztWqVYuq71tKy6PhzNWrV7d8+fK5M8r169fPTp06ZRnts88+szvvvNOdYU7fmfnz5yf5nE8//dTq16/v/i9VqlSxadOmWbRIaXnefvttu+2226xkyZLu5D06J81HH31k0SI1/x/PihUrLGfOnG6UXCyX5/Tp0/bEE09YpUqV3HdOvztNepjRMn1I8Kal1nCTDRs2WN26dd201IcOHfJdfuXKlda5c2fr3r27ffnll24HpMvXX39tsVgebehUnqVLl7oJtLThbt26tf34448Wi+XxfPfddzZgwAAXgKJFSsuiyc604VZZ3nrrLTdVukLdb37zG4vF8syaNcvN4qrlt27dalOmTHGv8fjjj1tG03lYtP4KPcmxe/dua9u2rd166622ceNG69u3r91///1Rs2NNaXm009J3TQFUE+ypXNqJaRsXi+UJDdn33XeftWrVyqLJiVSU56677nInFdTvRtuCN954wwXuDBfI5G688cZAr169grfPnz8fKFeuXCAhIcF3+bvuuivQtm3bsPsaNWoUePDBBwOxWJ7Ezp07FyhUqFBg+vTpgVgtj8pw0003BV577bVAt27dAu3atQvEYlkmTZoUuPbaawNnzpwJRKOUlkfLtmzZMuy+/v37B5o2bRqIJtrszZs377LLDBo0KHDdddeF3depU6dAfHx8INokpzx+atWqFRgxYkQglsuj/8nQoUMDTz31VKBu3bqBaGTJKM+HH34YKFKkSODnn38ORJtMXZPgTUutKvbkTkut+0OXFx09XWr5aC9PYidPnrSzZ89a8eLFLVbLM3LkSCtVqpSr7YkWqSnLu+++66p91dygE4ldf/319txzz7lZUmOxPDfddJN7jtcksWvXLnfkescdd1isiebtQCToVPia4CcatgOpNXXqVPcdU81VrHv33XfdGYhHjx7tahLV7Kia0n//+98ZvWqxMQtkek5LrTNA+i2v+2OxPIkNHjzYtZMl3gDGSnmWL1/uquNUBRxNUlMWbeA++eQT69Kli9uZ7tixwx5++GEX4jJ6w5ea8txzzz3uec2aNXOzzJ07d84eeuihqGhuSKlLbQc0e5823OpzEcteeOEFO378uKvijkXbt293TVuff/65648Q63bt2uW2ber7M2/ePPc70rbg559/dmEoI2XqmgSEGzVqlOvspy+hvoyxRkc+Xbt2de32mk00MxzNqUbklVdecZOe6XTl6rg0efJki0Xq/6KakIkTJ7o+DOos9/7779vTTz+d0auGRH1HRowYYXPmzHHfv1ij8KpAqjLoiDszuHDhguvgOHPmTDefkWrfxowZY9OnT8/w2oTYj2ARnpZa96dk+WgvT+iRg0LCxx9/bHXq1LFokNLy7Ny503XyU4er0B+X6GhCnX0qV65ssfK/0YiGXLlyued5atas6Y5iVd2vGVMzSmrK8+STT7oQpw5+opFB6sDVo0cPF37UXBErLrUd0MiAWK5F0EGC/j8awRUNtYmpPVhYt26d63TZu3fv4HZAtVfaDixatMhatmxpsaRs2bKumSF0+mZtC1SmH374wapWrZph6xY7v9p0mpZa94cuL4sXL46KaaxTO8222rl0NKdJtNTuFS1SWh4NS920aZNravAuv//974M90DVyI5b+N02bNnVNDF7QkW+//dZtMDIyIKS2POrvkjgIeAEo1qaIiebtQGqpt7xm2tW1Rm7EKgW1xNsBNWtpJID+1lDdWNO0aVPbt2+fawIK3Rbo91S+fPkMXbdMP7ph9uzZgTx58gSmTZsW2LJlS6BHjx6BokWLBg4cOOAe79q1a+Cxxx4LLr9ixYpAzpw5Ay+88EJg69atrtdsrly5Aps2bQrEYnlGjRoVyJ07d+Ctt94K7N+/P3g5duxYIBbLk1g0jW5IaVn27NnjRpr07t07sG3btsCCBQsCpUqVCjzzzDMZWIrUl0e/FZXnjTfeCOzatSuwaNGiQOXKld2IoYym7/uXX37pLtrsjRkzxv39/fffu8dVDpXHo/XPnz9/YODAgW47MGHChECOHDkCCxcuDESDlJZn5syZbrumcoRuBw4fPhyIxfIkFm2jG46lsDxavnz58oE//OEPgc2bNweWLVsWqFq1auD+++8PZLRMHxJk/PjxgYoVK7qdpYZ1rV69OvjYLbfc4nY0oebMmROoVq2aW17DoN5///1ArJanUqVK7kua+KIfVaz+f6I1JKSmLCtXrnRDbLUz1nDIZ5991g3xjMXynD17NjB8+HAXDPLmzRuoUKFC4OGHHw78+uuvgYy2dOlS39+Bt/66VnkSP6devXqu7PrfTJ06NRAtUloe/X255WPx/xPNIWFpKsqjMBoXFxfIly+fCwwaPnzy5MlARmOqaAAAkPX6JAAAgNQjJAAAAF+EBAAA4IuQAAAAfBESAACAL0ICAADwRUgAAAC+CAkAAMAXIQFAxFx99dU2bty4ZC+vCbs0+120Tf0N4P8REgBEzNq1a92sj5E0bdo0K1q0aERfE0DyZOqpogGkr5IlS2b0KgCIIGoSgCxswYIF7ij9/Pnz7raq/VX9/9hjjwWXuf/+++3ee+91fy9fvtyaN29u+fLlc1NzP/LII3bixIlLNjd888031qxZM8ubN6/VqlXLPv74Y/f68+fPD1uPXbt2uSm/8+fPb3Xr1rVVq1a5+z/99FM3vfGRI0fc83QZPnx4mn8uAP4fIQHIwrTDP3bsmH355Zfu9rJly+yqq65yO2eP7mvRooXt3LnTbr/9duvYsaN99dVX9uabb7rQ0Lt3b9/XVvBo37692/GvWbPGXnnlFXviiSd8l9X9AwYMcCGlWrVq1rlzZzt37pzddNNNLnQULlzY9u/f7y5aDkD6ICQAWViRIkWsXr16wVCg6379+rnQcPz4cfvxxx9tx44ddsstt1hCQoJ16dLF+vbta1WrVnU78JdfftlmzJhhp06duui1Fy9e7IKFHlftgGoUnn32Wd/10I6/bdu2LiCMGDHCvv/+e/e+uXPnduuoGoQyZcq4S8GCBdP8cwHw/wgJQBanAKBwoFnjP//8c+vQoYPVrFnT1RKoFqFcuXIuFPzzn/90nQi1k/Yu8fHxduHCBdu9e/dFr7tt2zbXJKEdu+fGG2/0XYc6deoE/y5btqy7PnToUJqUF0Dy0XERyOLUlPD666+7EJArVy6rUaOGu0/B4ddff3UhQlSz8OCDD7p+CIlVrFjxitZB7+tRrYEofADIWIQEIIvz+iWMHTs2GAgUEkaNGuVCwl/+8hd3X/369W3Lli1WpUqVZL1u9erVbe/evXbw4EErXbp0cIhkSqnJwetYCSB90dwAZHHFihVz1f0zZ8504UBuvvlm27Bhg3377bfB4DB48GBbuXKl66ioDobbt2+3d95555IdF2+77TarXLmydevWzXV0XLFihQ0dOjSstiA5NGJCtRhLliyxf/3rX3by5MmIlBtA0ggJAFwQ0NG6FxKKFy/uhiyqP4FqBERBQn0UFBxU+/Db3/7Whg0b5vos+MmRI4cb6qgd/A033OCGUnqjGzQkMrnUQfKhhx6yTp06ufMwjB49OiJlBpC0bAH1VgKAdKDaBI1y0MgF1TIAiG6EBABpZt68eW4UhEZHKBg8+uijrnlDIycARD86LgJIM+oQqb4Me/bscSdpiouLsxdffDGjVwtAMlGTAAAAfNFxEQAA+CIkAAAAX4QEAADgi5AAAAB8ERIAAIAvQgIAAPBFSAAAAL4ICQAAwPz8H2Z9lwvSxZ+KAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 600x350 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trimmed = IPTWWeighter(matching_data, trim_quantiles=(0.05, 0.95), verbose=True)\n",
    "trimmed.match()\n",
    "\n",
    "print(f\"Untrimmed: ESS={weighter.effective_sample_size():.1f}, max weight={weighter.weights.max():.2f}\")\n",
    "print(f\"Trimmed:   ESS={trimmed.effective_sample_size():.1f}, max weight={trimmed.weights.max():.2f}, n_trimmed={trimmed.diagnostics['n_trimmed']}\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(6, 3.5))\n",
    "ax.hist(weighter.weights, bins=30, alpha=0.6, label=\"untrimmed\")\n",
    "ax.hist(trimmed.weights, bins=30, alpha=0.6, label=\"trimmed (5th/95th pctile)\")\n",
    "ax.set_xlabel(\"weight\")\n",
    "ax.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ccf92cf",
   "metadata": {},
   "source": [
    "## Takeaways\n",
    "\n",
    "- `IPTWWeighter` fits a pool-vs-target propensity model and weights the pool by $\\hat p / (1 -\n",
    "  \\hat p)$; the target keeps weight 1 (the ATT estimand, target-as-reference).\n",
    "- It requires a **patient-level** target -- there's no aggregate-only (\"published Table 1\")\n",
    "  variant.\n",
    "- It does **not** solve for exact balance, so always check `weighted_balance_table()` and\n",
    "  `effective_sample_size()` on the fitted result, the same as for any other weighter.\n",
    "- `trim_quantiles` caps extreme weights, trading a little bias for a large reduction in variance --\n",
    "  worth reaching for whenever a handful of patients end up dominating the effective sample size."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pybal",
   "language": "python",
   "name": "pybal"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}