diff --git a/examples/2011-census-microdata/Example_Analysis.ipynb b/examples/2011-census-microdata/Example_Analysis.ipynb
new file mode 100644
index 0000000..b510d71
--- /dev/null
+++ b/examples/2011-census-microdata/Example_Analysis.ipynb
@@ -0,0 +1,1139 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "id": "native-desktop",
+ "metadata": {},
+ "source": [
+ "## Example classification task"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "focused-november",
+ "metadata": {},
+ "source": [
+ "This is a (work-in-progress) example of assessing synthetic data by comparing the performance of classifiers.\n",
+ "\n",
+ "For a given method, two classifiers are trained: one each on the original and synthetic data. These are both tested using a hold-out set from the *original* dataset. Similar performance is the desired outcome.\n",
+ "\n",
+ "This can be repeated for other classification methods and tasks.\n",
+ "\n",
+ "An alternative assessment is to compare the ranked performance of various methods (ignoring their absolute performance)."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "colored-clerk",
+ "metadata": {},
+ "source": [
+ "### Imports and setup"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "related-sharing",
+ "metadata": {},
+ "source": [
+ "The output from the `2011-census-test-*` examples must be present in the usual location (`../synth-output/`). "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "id": "spare-joyce",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from collections import namedtuple\n",
+ "import numpy as np\n",
+ "import pandas as pd\n",
+ "import seaborn as sns\n",
+ "import matplotlib.ticker as ticker\n",
+ "import matplotlib.pyplot as plt\n",
+ "\n",
+ "import sklearn\n",
+ "from sklearn.linear_model import LogisticRegression\n",
+ "from sklearn.ensemble import RandomForestClassifier\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "import sklearn.metrics as metrics\n",
+ "from sklearn.metrics import precision_score, recall_score, roc_auc_score, classification_report"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "surrounded-interval",
+ "metadata": {},
+ "source": [
+ "### Task"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "published-diesel",
+ "metadata": {},
+ "source": [
+ "Predict an individual's marital status from their other demographics.\n",
+ "\n",
+ "Simplification: Predict whether an individual is single."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "id": "juvenile-eight",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "features = [\"Sex\", \n",
+ " \"Age\",\n",
+ " \"Country of Birth\",\n",
+ " \"Health\",\n",
+ " \"Ethnic Group\",\n",
+ " \"Religion\",\n",
+ " \"Approximated Social Grade\",\n",
+ " \"Industry\"]"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "twenty-nomination",
+ "metadata": {},
+ "source": [
+ "### Load data"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "id": "treated-cancer",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "orig = pd.read_csv(\"../../datasets/2011-census-microdata/2011-census-microdata.csv\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "id": "indie-korea",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# synthetic datasets ('rlsd' = released)\n",
+ "rlsd = {}\n",
+ "rlsd[\"synthpop 3\"] = pd.read_csv(\"../../synth-output/2011-census-test-3-synthpop/synthetic_data_1.csv\")\n",
+ "rlsd[\"synthpop 5\"] = pd.read_csv(\"../../synth-output/2011-census-test-5-synthpop-cart/synthetic_data_1.csv\")\n",
+ "rlsd[\"synthpop 10\"] = pd.read_csv(\"../../synth-output/2011-census-test-10-synthpop-cart-proper/synthetic_data_1.csv\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "id": "rapid-cradle",
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "
\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " Person ID | \n",
+ " Region | \n",
+ " Residence Type | \n",
+ " Family Composition | \n",
+ " Population Base | \n",
+ " Sex | \n",
+ " Age | \n",
+ " Marital Status | \n",
+ " Student | \n",
+ " Country of Birth | \n",
+ " Health | \n",
+ " Ethnic Group | \n",
+ " Religion | \n",
+ " Economic Activity | \n",
+ " Occupation | \n",
+ " Industry | \n",
+ " Hours worked per week | \n",
+ " Approximated Social Grade | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 0 | \n",
+ " 7394816 | \n",
+ " E12000001 | \n",
+ " H | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 6 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 5 | \n",
+ " 8 | \n",
+ " 2 | \n",
+ " -9 | \n",
+ " 4 | \n",
+ "
\n",
+ " \n",
+ " 1 | \n",
+ " 7394745 | \n",
+ " E12000001 | \n",
+ " H | \n",
+ " 5 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 4 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 8 | \n",
+ " 6 | \n",
+ " 4 | \n",
+ " 3 | \n",
+ "
\n",
+ " \n",
+ " 2 | \n",
+ " 7395066 | \n",
+ " E12000001 | \n",
+ " H | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 4 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 6 | \n",
+ " 11 | \n",
+ " 3 | \n",
+ " 4 | \n",
+ "
\n",
+ " \n",
+ " 3 | \n",
+ " 7395329 | \n",
+ " E12000001 | \n",
+ " H | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 7 | \n",
+ " 7 | \n",
+ " 3 | \n",
+ " 2 | \n",
+ "
\n",
+ " \n",
+ " 4 | \n",
+ " 7394712 | \n",
+ " E12000001 | \n",
+ " H | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 5 | \n",
+ " 4 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 4 | \n",
+ " 3 | \n",
+ " 2 | \n",
+ "
\n",
+ " \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ " ... | \n",
+ "
\n",
+ " \n",
+ " 569736 | \n",
+ " 7946020 | \n",
+ " W92000004 | \n",
+ " H | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 5 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 4 | \n",
+ " 1 | \n",
+ " 9 | \n",
+ " 1 | \n",
+ " 8 | \n",
+ " 8 | \n",
+ " 3 | \n",
+ " 3 | \n",
+ "
\n",
+ " \n",
+ " 569737 | \n",
+ " 7944310 | \n",
+ " W92000004 | \n",
+ " H | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 7 | \n",
+ " 4 | \n",
+ " 3 | \n",
+ " 4 | \n",
+ "
\n",
+ " \n",
+ " 569738 | \n",
+ " 7945374 | \n",
+ " W92000004 | \n",
+ " H | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " -9 | \n",
+ " -9 | \n",
+ " -9 | \n",
+ " -9 | \n",
+ " -9 | \n",
+ "
\n",
+ " \n",
+ " 569739 | \n",
+ " 7944768 | \n",
+ " W92000004 | \n",
+ " H | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 8 | \n",
+ " 5 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 3 | \n",
+ " 1 | \n",
+ " 9 | \n",
+ " 5 | \n",
+ " 9 | \n",
+ " 2 | \n",
+ " -9 | \n",
+ " 4 | \n",
+ "
\n",
+ " \n",
+ " 569740 | \n",
+ " 7944959 | \n",
+ " W92000004 | \n",
+ " H | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 2 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 1 | \n",
+ " 7 | \n",
+ " 4 | \n",
+ " 1 | \n",
+ " 4 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
569741 rows × 18 columns
\n",
+ "
"
+ ],
+ "text/plain": [
+ " Person ID Region Residence Type Family Composition \\\n",
+ "0 7394816 E12000001 H 2 \n",
+ "1 7394745 E12000001 H 5 \n",
+ "2 7395066 E12000001 H 3 \n",
+ "3 7395329 E12000001 H 3 \n",
+ "4 7394712 E12000001 H 3 \n",
+ "... ... ... ... ... \n",
+ "569736 7946020 W92000004 H 1 \n",
+ "569737 7944310 W92000004 H 3 \n",
+ "569738 7945374 W92000004 H 3 \n",
+ "569739 7944768 W92000004 H 1 \n",
+ "569740 7944959 W92000004 H 2 \n",
+ "\n",
+ " Population Base Sex Age Marital Status Student Country of Birth \\\n",
+ "0 1 2 6 2 2 1 \n",
+ "1 1 1 4 1 2 1 \n",
+ "2 1 2 4 1 2 1 \n",
+ "3 1 2 2 1 2 1 \n",
+ "4 1 1 5 4 2 1 \n",
+ "... ... ... ... ... ... ... \n",
+ "569736 1 1 5 1 2 1 \n",
+ "569737 1 1 3 1 2 1 \n",
+ "569738 1 1 1 1 1 1 \n",
+ "569739 1 2 8 5 2 1 \n",
+ "569740 1 2 2 2 2 1 \n",
+ "\n",
+ " Health Ethnic Group Religion Economic Activity Occupation \\\n",
+ "0 2 1 2 5 8 \n",
+ "1 1 1 2 1 8 \n",
+ "2 1 1 1 1 6 \n",
+ "3 2 1 2 1 7 \n",
+ "4 1 1 2 1 1 \n",
+ "... ... ... ... ... ... \n",
+ "569736 4 1 9 1 8 \n",
+ "569737 2 1 1 1 7 \n",
+ "569738 1 1 2 -9 -9 \n",
+ "569739 3 1 9 5 9 \n",
+ "569740 2 1 1 1 7 \n",
+ "\n",
+ " Industry Hours worked per week Approximated Social Grade \n",
+ "0 2 -9 4 \n",
+ "1 6 4 3 \n",
+ "2 11 3 4 \n",
+ "3 7 3 2 \n",
+ "4 4 3 2 \n",
+ "... ... ... ... \n",
+ "569736 8 3 3 \n",
+ "569737 4 3 4 \n",
+ "569738 -9 -9 -9 \n",
+ "569739 2 -9 4 \n",
+ "569740 4 1 4 \n",
+ "\n",
+ "[569741 rows x 18 columns]"
+ ]
+ },
+ "execution_count": 5,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "orig"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "royal-southeast",
+ "metadata": {},
+ "source": [
+ "### Basic quality checks"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "id": "dying-louisiana",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# features to visualize\n",
+ "features_subset = [\"Age\", \"Marital Status\", \"Religion\", \"Occupation\"]\n",
+ "\n",
+ "n_plot_sample = 200000\n",
+ "\n",
+ "# Kludge: Some visualizations are more straightforward with continuous data, so convert the factors to float\n",
+ "\n",
+ "def recode(df):\n",
+ " return df.applymap(\n",
+ " lambda x: float(x) - 1 if int(x) != -9 else np.NAN\n",
+ " )\n",
+ "\n",
+ "\n",
+ "\n",
+ "def format_axis_ticks(grid):\n",
+ " for ax in grid.axes.flat:\n",
+ " if ax is not None:\n",
+ " xticks = ax.get_xticks()\n",
+ " xticks_new = [str(int(t + 1)) for t in xticks]\n",
+ "\n",
+ " yticks = ax.get_yticks()\n",
+ " yticks_new = [str(int(t + 1)) for t in yticks]\n",
+ " \n",
+ " ax.set_xticklabels(xticks_new)\n",
+ " ax.set_yticklabels(yticks_new)\n",
+ "\n",
+ " return grid\n",
+ "\n",
+ "def jitter(x):\n",
+ " return x + 0.3 * np.random.randn()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "id": "abstract-charlotte",
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/var/folders/4t/76rq38xx7pn94vfcd3n5l2m0l8wv2w/T/ipykernel_10396/2062598768.py:25: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
+ " ax.set_yticklabels(yticks_new)\n"
+ ]
+ },
+ {
+ "data": {
+ "text/plain": [
+ ""
+ ]
+ },
+ "execution_count": 7,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAArEAAALBCAYAAABP+PnsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAA56UlEQVR4nO3debikeVkf/O/dy+wzDAwNooDIIqgoA9NMcAFZBAdEUCAiUVE0TF5FBE0m0byJW0xcIAmir8aJICCKshohZoQAggsBemCAYZOwKQhOD9vMMMzQy/3+UdXYjH3OqVPPqarz9Pl8rquurqpT9/O7zzm/qvqep3/1PNXdAQCAMdm16gYAAGCzhFgAAEZHiAUAYHSEWAAARkeIBQBgdPasuoGbcKgENqtW3QAAsHz2xAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjs2fVDWzGBZc8b1OPv/xpj19QJwAArJI9sQAAjI4QCwDA6AixAACMjhALAMDoCLEAAIyOEAsAwOgIsQAAjI4QCwDA6AixAACMjhALAMDoCLEAAIyOEAsAwOjsWXUDO8kFlzxv0zWXP+3xW1YPAHCyEGI3QYgEANgedlSIFUIBAE4O1sQCADA6QiwAAKOzo5YT7HSWUwAAJwt7YgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABgdIRYAgNERYgEAGB3HiWVmjjMLAGwX9sQCADA69sSyNPbkAgBbxZ5YAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdKq7V93DF1TVZUluOUfpLZNcPWDonV6/HXqYt/7q7r5owLgAwAhtqxA7r6o60N371c9v1T1sxfcAAOwclhMAADA6QiwAAKNzsoTYS9UPtuoetuJ7AAB2iJNiTSwAADvLybInFgCAHUSIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGZ8+qGzjeRRdd1Jdddtmq22BcaoljOb0d81jmHAXYMbbVntirr7561S0AADAC2yrEAgDALIRYAABGR4gFAGB0hFgAAEZHiAUAYHS21SG2FuUlb/rUqlsY5MjAAzsdOjLsCD/XHx5W//7rhk+zX3rYmYO3AQCcPOyJBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZnoaedraoPJbk2yZEkh7t7/yLHAwBgZ1hoiJ16QHdfvYRxAADYISwnAABgdBYdYjvJK6vq8qq6+EQPqKqLq+pAVR04ePDggtsBAOBksOgQ+03dfa8kD03ypKq6300f0N2Xdvf+7t6/b9++BbcDAMDJYKEhtrs/Ov33qiQvS3LhIscDAGBnWFiIraozq+rsY9eTPCTJlYsaDwCAnWORRye4dZKXVdWxcX6/uy9b4HgAAOwQCwux3f2BJPdY1PYBANi5HGILAIDREWIBABgdIRYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABidRZ6xa8u85E2fWnULMJMLLnnelm/z8qc9fsu3CQBjN4oQu2qHjw6r//Tnh+3wvu7QsPpTd/Wg+rP2DvsBnH+LQ4PqAQBuynICAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0NgyxVfUrVXVOVe2tqldX1cGq+t5lNAcAACcyy57Yh3T3NUkenuRDSe6c5JJZB6iq3VX11qp6xXwtAgDAF5slxB47Ne23JXlRd39mk2M8Jcm7N1kDAABrmiXEvqKq3pPkgiSvrqp9SW6YZeNVddtMwu9vz98iAAB8sQ1DbHf/ZJJvSLK/uw8l+WySR864/Wck+ddJjq71gKq6uKoOVNWBgwcPzrhZAAB2sj0bPaCqHn/c9eO/9LwN6h6e5Kruvryq7r/W47r70iSXJsn+/ft7o34AAGDDEJvk3sddPy3Jg5K8JRuE2CTfmOQRVfWwad05VfX87nZkAwAABtkwxHb3k4+/XVXnJvmDGep+KslPTWvun+RfCbAAAGyFeU528NkkX7HVjQAAwKxmWRP78iTH1qruSvLVSV60mUG6+8+S/NkmewMAgBOaZU3s04+7fjjJh7v7IwvqBwAANjTLcoKHdffrppe/7O6PVNUvL7wzAABYwywh9sEnuO+hW90IAADMas3lBFX1w0l+JMkdq+rtx33p7CR/uejGAABgLeutif39JP8ryS8m+cnj7r+2uz+50K4AAGAda4bY7v5Mks8keVySVNWtMjlpwVlVdVZ3/81yWgRu6oJLNjrXyOZd/rTHb/wgANgmNlwTW1XfXlXvS/LBJK9L8qFM9tACAMBKzPLBrl9Icp8kf93dX5HJaWf/z0K7AgCAdcwSYg919yeS7KqqXd392iT7F9wXAACsaZaTHXy6qs5K8vokv1dVV2Vy6tmlefSFNx9U/5I3fWqLOpnPWXt64wetW39kUP2eXcPG312DylMD6wEAbmqWPbGPTHJ9kh9PclmS9yd5+CKbAgCA9cwSYn+6u4929+Hufm53PzPJv1l0YwAAsBZn7AIAYHQ2e8auSnJWnLELAIAVcsYuAABGZ73lBIeSfLS7H9fdH87kbF2PSnL/ZTQGAABrWS/EXpbkDklSVXdO8oYkd0zypKr6pcW3BgAAJ7ZeiL15d79vev37k7ygu5+cyYe6vm3hnQEAwBrWC7HHHyH/gUlelSTd/fkkRxfZFAAArGe9D3a9vaqenuSjSe6c5JVJUlXnzrLhqjotk7N8nTod58Xd/TODugUAgKy/J/aJSa7OZF3sQ7r7+un9X53k6TNs+8YkD+zueyQ5P8lFVXWf+VsFAICJNffEdvfnkvyjD3B1918l+auNNtzdneS66c2900uvXQEAALOZ5Yxdc6uq3VV1RZKrkryqu994gsdcXFUHqurAwYMHF9kOAAAniYWG2O4+0t3nJ7ltkgur6u4neMyl3b2/u/fv27dvke0AAHCSWGiIPaa7P53ktUkuWsZ4AACc3NZcE1tVL886a1i7+xHrbbiq9iU51N2frqrTkzw4yS/P2ygAAByz3iG2ZjkCwXpuk+S5VbU7kz2+L+zuVwzcJgAArHt0gtcN2XB3vz3JPYdsAwAATmS9PbFJkqq6S5JfzOT4sKcdu7+777jAvgAAYE2zfLDrd5L8ZpLDSR6Q5HlJnr/IpgAAYD2zhNjTu/vVSaq7P9zdP5vk2xbbFgAArG3D5QRJbqyqXUneV1U/muSjSc5abFsAALC2WfbEPiXJGUl+LMkFSb43yeMX2RQAAKxnlhB7h+6+rrs/0t1P6O5HJ7n9ohsDAIC1zBJif2rG+wAAYCnWO2PXQ5M8LMmXVdUzj/vSOZkcqQAAAFZivQ92/V2SA0kekeTy4+6/NsmPL7Kpm3rJmz61zOEAANjm1jtj19uSvK2qfq+7R73n9fDRYfWv/Pipg+rf9InPDqo/euiGQfW1e++g+ludc+6g+gvPPjKoPkkePXgLAMDJZL3lBC/s7u9K8taq6pt+vbu/bqGdAQDAGtZbTvCU6b8PX0YjAAAwq/WWE3ysqnYneU53P2CJPQEAwLrWPcRWdx9JcrSqbrakfgAAYEOznHb2uiTvqKpXJfnCJ5S6+8cW1hUAAKxjlhD70ukFAAC2hQ1DbHc/dxmNAADArDYMsVV1lyS/mOSrk5x27P7uvuMC+wIAgDWt+8Guqd9J8puZnGr2AUmel+T5i2wKAADWM0uIPb27X52kuvvD3f2zSb5to6Kqul1Vvbaq3lVV76yqp2xUAwAAs5jlg103VtWuJO+rqh9N8tEkZ81QdzjJv+zut1TV2Ukur6pXdfe7BvQLAAAz7Yl9SpIzkvxYkguSfF+S79+oqLs/1t1vmV6/Nsm7k3zZ/K0CAMDELEcnePP06nVJnjDPIFV1hyT3TPLGE3zt4iQXJ8ntb3/7eTYPAMAOs2aIrao/Xq+wux8xywBVdVaSlyR5andfc4LtXJrk0iTZv39/z7JNAAB2tvX2xH59kr9N8oJM9qDWZjdeVXszCbC/191OmAAAwJZYL8R+SZIHJ3lckn+W5H8meUF3v3OWDVdVJXlWknd3938Z2igAAByz5ge7uvtId1/W3d+f5D5J/m+SP5seoWAW35jJh8AeWFVXTC8PG94yAAA73bof7KqqUzM5JuzjktwhyTOTvGyWDXf3X2SOJQgAALCR9T7Y9bwkd0/yJ0l+rruvXFpXAACwjvX2xH5vks9mcpzYH5sscU0y2bva3X3OgnsDAIATWjPEdvcsJ0IAAIClE1QBABgdIRYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDRWfeMXdvFrz7tZwfVX3+3hw6q/9yX3X1Q/X1vc4tB9Xc567RB9eedemRQ/am7bxhUX87bBgBsMXtiAQAYHSEWAIDREWIBABgdIRYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABidhYXYqnp2VV1VVVcuagwAAHamRe6JfU6Sixa4fQAAdqiFhdjufn2STy5q+wAA7FwrXxNbVRdX1YGqOnDw4MFVtwMAwAisPMR296Xdvb+79+/bt2/V7QAAMAIrD7EAALBZe1bdAMAFlzxvS7d3+dMev6XbA2D7WeQhtl6Q5A1J7lpVH6mqH1rUWAAA7CwL2xPb3Y9b1LYBANjZrIkFAGB0rIkFdoStXnebWHsLsEr2xAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDqjOE7s61/0q6tuAQCAbWQUIRZYja0+QYCTAwCwVSwnAABgdIRYAABGR4gFAGB0hFgAAEanunvVPXxBVR1M8uE5Sm+Z5OoBQ+/0+u3Qw7z1V3f3RQPGnVlVXZZJn1tlK35v22EM46xvaXMUYCfZViF2XlV1oLv3q5/fqnvYiu9hbJbxPS/r52ocAJbNcgIAAEZHiAUAYHROlhB7qfrBVt3DVnwPY7OM73lZP1fjALBUJ8WaWAAAdpaTZU8sAAA7iBALAMDoCLEAAIyOEAsAwOgIsQAAjI4QCwDA6AixAACMjhALAMDoCLEAAIzOtgqxF110USdxcdnMZWnMT5c5L8u06u/VZZwXGKVtFWKvvvrqVbcAazI/AWD72FYhFgAAZiHEAgAwOkIsAACjI8QCADA6e1bdwCx+/k+vG1R/y1OPDqpf9Uc3Dw1rP3tqWP2pu1f9E0ieeL+brbqFNf3HV167lHHOPWU5v4dbn35k4WMcOjpwUs5oVy1v7j72Pjdf2lgAjCTEArA6F1zyvE09/vKnPX5BnQD8A8sJAAAYHSEWAIDREWIBABgdIRYAgNFZaIitqqdU1ZVV9c6qeuoixwIAYOdYWIitqrsneWKSC5PcI8nDq+rOixoPAICdY5F7Yr8qyRu7+/ruPpzkdUketcDxAADYIRYZYq9Mct+qOq+qzkjysCS3u+mDquriqjpQVQcOHjy4wHZg88xPANieFhZiu/vdSX45ySuTXJbkiiT/6FRA3X1pd+/v7v379u1bVDswF/MTALanhX6wq7uf1d0XdPf9knwqyV8vcjwAAHaGhZ52tqpu1d1XVdXtM1kPe59FjgcAwM6w0BCb5CVVdV6SQ0me1N2fXvB4AADsAAsNsd1930VuHwCAnckZuwAAGB0hFgCA0Vn0mtgt8dXnHlp1CyvVA+traP3QDZzk7nqzw6tuYUst4/d9yu6hsxqAnc6eWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGZ8+qG5jFkR5Wv6uG1R9d8fhDDWx/8AZqxd//oh1ddQNbbPeqGwCAGdgTCwDA6AixAACMjhALAMDoLCzEVtVdq+qK4y7XVNVTFzUeAAA7x8I+2NXd701yfpJU1e4kH03yskWNBwDAzrGs5QQPSvL+7v7wksYDAOAktqwQ+91JXnCiL1TVxVV1oKoOHDx4cEntwGzMTwDYnhYeYqvqlCSPSPKiE329uy/t7v3dvX/fvn2Lbgc2xfwEgO1pGXtiH5rkLd3990sYCwCAHWAZIfZxWWMpAQAAzGOhIbaqzkzy4CQvXeQ4AADsLAs7xFaSdPdnk5y3yDEAANh5nLELAIDREWIBABidhS4n2Crv+NTeQfW3O/PIoPobjtSg+lN39aD6M/cMq9+7e1B5dtew8TOwfLv72PUDf8AzOn33cn6Qtzj16FLGWYYa9tQFYBuzJxYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABidmUJsVT2qqt5XVZ+pqmuq6tqqumbRzQEAwInMepzYX0ny7d397kU2AwAAs5h1OcHfC7AAAGwXs+6JPVBVf5jkj5LceOzO7n7pIpoCAID1zBpiz0lyfZKHHHdfJxFiAQBYuplCbHc/YdGNAADArGY9OsFtq+plVXXV9PKSqrrtopsDAIATmfWDXb+T5I+TfOn08vLpfQAAsHSzhth93f073X14enlOkn0L7AsAANY06we7PlFV35vkBdPbj0vyicW09I/d4xaHBtXXFvWxU9XAH2D31vSxXX3pGUeWMs6y5vHQ3zcALMOse2J/MMl3Jfl4ko8leUwSH/YCAGAlZj06wYeTPGLBvQAAwEzWDbFV9a+7+1eq6tcyOS7sF+nuH1tYZwAAsIaN9sQeO9XsgXk2XlUfSnJtkiNJDnf3/nm2AwAAx1s3xHb3y6f/PnfAGA/o7qsH1AMAwBeZaU1sVb08/3g5wWcy2UP7W919w1Y3BgAAa5n16AQfSHJdkv8+vVyTyTKBr5zeXksneWVVXV5VF5/oAVV1cVUdqKoDBw8enL1zWALzEwC2p1mPE/sN3X3v426/vKre3N33rqp3rlP3Td390aq6VZJXVdV7uvv1xz+guy9NcmmS7N+//yQ/oihjY34CwPY0657Ys6rq9sduTK+fNb35+bWKuvuj03+vSvKyJBfO2ScAAHzBrHti/2WSv6iq92dy4qCvSPIjVXVmkhN+6Gv6tV3dfe30+kOS/PwW9AwAwA4368kO/qSq7pLkbtO73nvch7mesUbZrZO8rCbnsNyT5Pe7+7IBvQIAQJKNT3bwwO5+TVU96iZfulNVpbtfulZtd38gyT22okkAADjeRntivznJa5J8+wm+1knWDLEAALAoG53s4Gem/z5hOe0AAMDGNlpO8BPrfb27/8vWtnNih4/WoPoaVp7DR4fV7x44/tD+d9WwI0PtcmCpdR3pgb+gGR1d0u/h1N1+4QBsfxstJzh7KV0AAMAmbLSc4OeW1QgAAMxqppMdVNVXVtWrq+rK6e2vq6p/t9jWAADgxGY9Y9d/T/JTSQ4lSXe/Pcl3L6opAABYz6wh9ozuftNN7ju81c0AAMAsZj3t7NVVdadMjg2bqnpMko8trCsAttwFlzxv5sde/rTHL7ATgOFmDbFPSnJpkrtV1UeTfDDJ9yysKwAAWMdMIXZ6CtlvqaozM1mCcH0ma2I/vMDeAADghNZdE1tV51TVT1XVr1fVgzMJr9+f5P8m+a5lNAgAADe10Z7Y303yqSRvSPLEJP9vkkrynd19xWJbAwCAE9soxN6xu782SarqtzP5MNftu/uGhXcGAABr2OgQW4eOXenuI0k+IsACALBqG+2JvUdVXTO9XklOn96uJN3d5yy0OwAAOIF1Q2x3715WIwAAMKtZz9gFAADbxqwnO1ipvbt6pePvXvH+6Frt8KlVN7DN7aklzU+/BwD4AntiAQAYHSEWAIDREWIBABidhYXYqjqtqt5UVW+rqndW1c8taiwAAHaWRX6w68YkD+zu66pqb5K/qKr/1d3/Z4FjAgCwAywsxHZ3J7luenPv9LLawwwAAHBSWOia2KraXVVXJLkqyau6+40neMzFVXWgqg4cPHhwke3AppmfALA9LTTEdveR7j4/yW2TXFhVdz/BYy7t7v3dvX/fvn2LbAc2zfwEgO1pKUcn6O5PJ3ltkouWMR4AACe3RR6dYF9VnTu9fnqSByd5z6LGAwBg51jk0Qluk+S5VbU7k7D8wu5+xQLHAwBgh1jk0QnenuSei9o+AAA7lzN2AQAwOkIsAACjs8g1sVvmmU//2UH115z/XYPq99xw7aD6Uz/5wUH1uz/zkUH1OfXsQeU33vLOg+qP7jltUH2SPPrChw/exqI88zd+dTkDXf+ppQzzQ09d/Bmi3/WZvQsfI0l2L2WUiUcvcSwA7IkFAGCEhFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdIRYAABGR4gFAGB0hFgAAEZHiAUAYHT2rLqBWVxzj386qP7QLW43rP7znxtU/7kvu/ug+tPPOW9Q/X3POTqo/s5nHRlUf+aeYeNvd9fe6QFLGecrv+bCpYxz6u4bFz7Gnc86vPAxkuScU5Y5985c4lgA2BMLAMDoCLEAAIyOEAsAwOgsbE1sVT07ycOTXNXdwxaFAgBJkgsued6mHn/50x6/oE5gtRa5J/Y5SS5a4PYBANihFhZiu/v1ST65qO0DALBzrXxNbFVdXFUHqurAwYMHV90OfBHzEwC2p5WH2O6+tLv3d/f+ffv2rbod+CLmJwBsTysPsQAAsFmjOGMXACyST/zD+CxsT2xVvSDJG5Lctao+UlU/tKixAADYWRa2J7a7H7eobQMAsLNZEwsAwOgIsQAAjM4oPth1xb/7plW3AGt6609/86pb2GKnr7oBANiQPbEAAIyOEAsAwOiMYjkBADvLZo7b6pitsDPZEwsAwOjYEwsAA9hrDKshxAKwEKs4leuYAuWYeoXtqLp71T18QVUdTPLhOUpvmeTqAUPv9Prt0MO89Vd390UDxp3ZgPm5lq34vW2HMYyzvmXO0csy6X2z5v1+x1K3ijHH1OvS5ihspW0VYudVVQe6e7/6+a26h634HsZmGd/zsn6uxhm3eb/fsdStYswx9Qpj5YNdAACMjhALAMDonCwh9lL1g626h634HsZmGd/zsn6uxhm3eb/fsdStYswx9QqjdFKsiQUAYGc5WfbEAgCwgwixAACMjhALAMDoCLEAAIyOEAsAwOgIsQAAjI4QCwDA6AixAACMjhALAMDobKsQe9FFF3USF5fNXJbG/HSZ87I05qjLnJdlWvX36jK+y5q2VYi9+uqrV90CrMn8ZLszR4GdZFuFWAAAmIUQCwDA6AixAACMjhALAMDo7Fl1A8vwjT/224PqP3eLOwyqP7r3jEH1uw7fMGz8PacNqj901nmD6rNr97D6JO988p0Hb2NR7vVzr1nKOF3Df46zeOqD7rHwMT5w7XJeeqqWMkyS5Ge/9azlDbZJ5z/vmrnqzvmfPzNX3ZN/4qfnqkuSIz3fL23vrnU/xLymaz4//76cM/cenatu3hE/8tlhrwFPfeA5g+phu7EnFgCA0RFiAQAYHSEWAIDREWIBABgdIRYAgNFZaIitqqdU1ZVV9c6qeuoixwIAYOdYWIitqrsneWKSC5PcI8nDq2r7HicJAIDRWOSe2K9K8sbuvr67Dyd5XZJHLXA8AAB2iEWG2CuT3LeqzquqM5I8LMntbvqgqrq4qg5U1YGDBw8usB3YPPOT7c4cBXaqhYXY7n53kl9O8soklyW5IsmREzzu0u7e39379+3bt6h2YC7mJ9udOQrsVAv9YFd3P6u7L+ju+yX5VJK/XuR4AADsDAs9gXlV3aq7r6qq22eyHvY+ixwPAICdYaEhNslLquq8JIeSPKm7P73g8QAA2AEWGmK7+76L3D4AADuTM3YBADA6QiwAAKOz6DWx28KjHvnYQfUvvGrvoPrrr792UP2hz356UH1OOX1Q+Z7Tzhw2/KnDxt/uLjx/OZ9X/PMP/s1Sxtm7hD9tb37K0cUPkuS03b2Ucba7B5+3e666V97ju+aq21VzlU0t93e2p+Yfb95vs+YsPHOP+QzHsycWAIDREWIBABgdIRYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABgdIRYAgNERYgEAGB0hFgCA0RFiAQAYHSEWAIDREWIBABidPatuYBle+kcvGFR/3d0ePKh+z5nnDqu/+ZcMqj9lz95B9Xc+tQfVn7vn6KD67e7P/+bvljLO7pvtW8o4y3Dq7mFzalZVSxlm2/u7G+fbX7Hr8OfnqusBv96jc9bumvN3vZyZeJMxVzEonITsiQUAYHSEWAAARkeIBQBgdIRYAABGZ2EhtqruWlVXHHe5pqqeuqjxAADYORZ2dILufm+S85OkqnYn+WiSly1qPAAAdo5lLSd4UJL3d/eHlzQeAAAnsWWF2O9OcsKDtVbVxVV1oKoOHDx4cEntwGzMT7Y7cxTYqRYeYqvqlCSPSPKiE329uy/t7v3dvX/fvpPnYO6cHMxPtjtzFNiplrEn9qFJ3tLdf7+EsQAA2AGWEWIflzWWEgAAwDwWGmKr6swkD07y0kWOAwDAzrKwQ2wlSXd/Nsl5ixwDAICdZ6YQW1VfmeSSJF9+fE13P3BBfQEAwJpm3RP7oiT/Lcl/T3Jkce0AAMDGZg2xh7v7NxfayQJde89HD9vA9Z8ZVH7kUx8bVn/08KD6Q+fcalD93tNuNqj+NqcdHVS/3Z12i9ssZZwbPvXxpYxz5OjiVwBdf3g5h6g+95STe+7N6szdPVfd7kPXb3EnM4xZyx1v74CpuGvOXnu+X8fc48HJatan78ur6keq6jZVdYtjl4V2BgAAa5h1T+z3T/+95Lj7Oskdt7YdAADY2Ewhtru/YtGNAADArGY9OsHeJD+c5H7Tu/4syW9196EF9QUAAGuadTnBbybZm+Q3pre/b3rfP19EUwAAsJ5ZQ+y9u/sex91+TVW9bRENAQDARmY9OsGRqrrTsRtVdcc4XiwAACsy657YS5K8tqo+kKQyOXPXExbWFQAArGPWoxO8uqrukuSu07ve2903Lq4tAABY27ohtqoe2N2vqapH3eRLd66qdPdLF9gbAACc0EZ7Yr85yWuSfPsJvtZJhFgAAJZu3RDb3T8zvfrz3f3B479WVU6AAADASsx6dIKXnOC+F29lIwAAMKuN1sTeLcnXJLnZTdbFnpPktEU2BgAAa9loTexdkzw8ybn54nWx1yZ54oJ6AgCAdW20JvZ/JPkfVfX13f2GJfW05U4//cxB9ddde/Wg+t2f+8yg+j03Xjuo/vNHh52X4s2Hhh1N7cDe4Tvt/13OGryNRfn8DdcvZZwzPvbupYyz617ftPAxDvfCh0iSHDq6nHG2uys/N+vKsS922ic+MOeI95mzbvmGTMVe0jw+5oYjtdwBYZub9WQHb62qJ2WytOALiaS7f3AhXQEAwDpm/fP8d5N8SZJvTfK6JLfNZEkBAAAs3awh9s7d/e+TfLa7n5vk25L8k8W1BQAAa5t1OcGh6b+frqq7J/l4klttVFRVH8pkj+2RJIe7e/88TQIAwPFmDbGXVtXNk/z7JH+c5Kzp9Vk8oLuHfTIKAACOM1OI7e7fnl59XZI7Lq4dAADY2ExrYqvqvKr6tap6S1VdXlXPqKrzZijtJK+c1ly8xrYvrqoDVXXg4MGDm+kdFs78ZLszR4GdatYPdv1BkquSPDrJY5JcneQPZ6j7pu6+V5KHJnlSVd3vpg/o7ku7e39379+3b9+M7cBymJ9sd+YosFPNGmJv093/obs/OL38QpJbb1TU3R+d/ntVkpcluXD+VgEAYGLWEPvKqvruqto1vXxXkj9dr6Cqzqyqs49dT/KQJFcOaxcAAGY/OsETkzw1yfOnt3cl+WxV/Ysk3d3nnKDm1kleVlXHxvn97r5sWLsAADD70QnO3uyGu/sDSe6x6Y4AAGADM4XYE30gK0m6+/Vb2w4AAGxs1uUElxx3/bRMPqB1eZIHbnlHAACwgVmXE3z78ber6nZJnrGIhhbhaA+r3/X56wfV19Ejw+qPHNr4Qes49ZqPD6r/3ClnDKrPDdcNq08yw8EwVmey7nvh9tzwmaWMswyHjy5nnKtv3L2cgba5W8+6u+ImPvHJD21pH7M4Mufr9by/6eU8e7fG4YHvZXCymfXoBDf1kSRftZWNAADArGZdE/trmZx9K5kE3/OTvGVBPQEAwLpm/U+mA8ddP5zkBd39lwvoBwAANjRriH1xkhu6+0iSVNXuqjqju4ctFgUAgDnMuib21UlOP+726Un+99a3AwAAG5s1xJ7W3V/4iPn0+sCPrAMAwHxmDbGfrap7HbtRVRck+dxiWgIAgPXNuib2qUleVFV/l8lh9b4kyWMX1RQAAKxn1pMdvLmq7pbkrtO73tvdw47ADwAAc5ppOUFVPSnJmd19ZXdfmeSsqvqRxbYGAAAnNuua2Cd296eP3ejuTyV54kI6AgCADcwaYndX/cMJ4qtqd5JTFtMSAACsb9YPdv1pkj+sqt+a3v5/kly2mJYAAGB9s4bYf5/J8oFj62D/NMmzFtIRAABsYN0QW1V7kvynJE9I8rfTu2+f5AOZLEU4stDuAADgBDbaE/u0JGcnuWN3X5skVXV2kv+c5OlJnrLY9rbGabOu/F3DKX/7lkH1h087Z1D9DefedlD9kVvdcVD9vc8b1v9tTzu5/9Y5euP1Sxnncze//VLGWYYz9/RSxrnlST73ZvXhGw7PVbf7Tt+8xZ0sznJm1NaYt9c9tfFjYCfZKN49PJMjE1x77I7p9R9O8rBFNgYAAGvZKMR2d/+jPxq7+0jG9YcvAAAnkY1C7Luq6vE3vbOqvjfJexbTEgAArG+jNbFPSvLSqvrBJJdP79uf5PQk37leYVWdluT1SU6djvPi7v6ZYe0CAMAGIba7P5rkn1TVA5N8zfTuP+nuV8+w7RuTPLC7r6uqvUn+oqr+V3f/n2EtAwCw0810nNjufk2S12xmw9O1tNdNb+6dXqyjBQBgsIEHn1pfVe2uqiuSXJXkVd39xhM85uKqOlBVBw4ePLjIdmDTzE+2O3MU2KkWGmK7+0h3n5/ktkkurKq7n+Axl3b3/u7ev2/fvkW2A5tmfrLdmaPATrXQEHtMd386yWuTXLSM8QAAOLktLMRW1b6qOnd6/fQkD47DcgEAsAVm+mDXnG6T5LlVtTuTsPzC7n7FAscDAGCHWFiI7e63J7nnorYPAMDOtZQ1sQAAsJWEWAAARmeRa2K3jXueeXRQ/RUfe/ug+t2DqpNTbnbbQfWHPvU3g+rffKdvHFT/tjNvNqg+SX5h8BYW6NANSxnmnLe9eCnj9AO/duFjfO5ILXyMJLn20HLG2e5uvPaTc9Wd99evmquuM/9rxqGj8/3OTtk937l0Dg14e9iz5N1ANyzpeQNjYU8sAACjI8QCADA6QiwAAKMjxAIAMDpCLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKOzI87YBQCcHC645Hmbrrn8aY9fQCesmj2xAACMjj2xAAAs1CL2oNsTCwDA6NgTCwAwA+txtxd7YgEAGJ0dsSf2K886PKj+NQ/5t4Pq9/7duwbV7zo6rP8bbvkVg+rPutktB9Xf9bQeVL/d3ffLb7uUcQ5c9a1LGWdXLX6MM/YsZ06ctnspw2x797z1vrnq3v1VD5urbsgU2rtrvrkx75h7V7ArZ95ez9l7dEv7gLHbESEWAGDVLEfYWkIsALBjbDZICpHblzWxAACMzsL2xFbVs5M8PMlV3X33RY0DAIyH/1JnqyxyOcFzkvx6ks3PVgAAvsjQpRAn21KKhS0n6O7XJ/nkorYPAMDOtfIPdlXVxUkuTpLb3/72K+4Gvpj5yXZnjgLLsB2Xgaz8g13dfWl37+/u/fv2zXcsQ1gU85PtzhwFdqqVh1gAANgsIRYAgNFZWIitqhckeUOSu1bVR6rqhxY1FgAAO8vCPtjV3Y9b1LYBANjZLCcAAGB0hFgAAEZn5ceJXYYffcA5K61Pbj2wnu3svz3yjOUM9MiHLWccTjrPf9Tp8xU+6lu3thGALWRPLAAAoyPEAgAwOkIsAACjI8QCADA6QiwAAKMjxAIAMDrV3avu4Quq6mCSD89ResskVw8YeqfXb4ce5q2/ursvGjDuzAbMz7Vsxe9tO4xhnPWNYY7O+/2OpW4VY46p12XO0csy6fNExvo+tBX1Y+59aP1GtWvOz20VYudVVQe6e7/6+a26h634HsZmGd/zsn6uxhm3eb/fsdStYswx9bpdrPp9ZJX1Y+59aP2QWssJAAAYHSEWAIDROVlC7KXqB1t1D1vxPYzNMr7nZf1cjTNu836/Y6lbxZhj6nW7WPX7yCrrx9z70Pq5a0+KNbEAAOwsJ8ueWAAAdhAhFgCA0Rl1iK2qZ1fVVVV15Zz1t6uq11bVu6rqnVX1lE3Wn1ZVb6qqt03rf27OPnZX1Vur6hVz1H6oqt5RVVdU1YE56s+tqhdX1Xuq6t1V9fWbqL3rdNxjl2uq6qmbHP/Hpz+7K6vqBVV12ma/h7EZOm9nHGPQ3N7EOFvyHJhxrLmfJ5sYY9DzaUzmnYdD5tbQ+TLPHBjyO53n9XHI6+K8r4dV9ZRpzTs3+xq8HZwM7+U79X186Hv44Lnb3aO9JLlfknsluXLO+tskudf0+tlJ/jrJV2+ivpKcNb2+N8kbk9xnjj5+IsnvJ3nFHLUfSnLLAT/D5yb559PrpyQ5d87t7E7y8SRfvomaL0vywSSnT2+/MMkPrHpeLfoydN7OOMagub2JcbbkOTDjWHM/TzYxxqDn05gu887DIXNr6HyZZw4M+Z0OfX3czOvivK+HSe6e5MokZyTZk+R/J7nzqufXMubicfUrfy/fie/jQ9/Dt2LujnpPbHe/PsknB9R/rLvfMr1+bZJ3Z/JLmbW+u/u66c2908umPilXVbdN8m1JfnszdVuhqm6WyYvHs5Kkuz/f3Z+ec3MPSvL+7t7s2YL2JDm9qvZkMpH/bs7xR2PovJ1xjEFzexPjDH4OzGKVz5OT1bzzcMjcGjJflj0Htuj1cbOvi/O8Hn5Vkjd29/XdfTjJ65I8apN9rtTY38t3+Pv4kPfwwXN31CF2K1XVHZLcM5O/wDZTt7uqrkhyVZJXdfem6pM8I8m/TnJ0k3XHdJJXVtXlVXXxJmu/IsnBJL8z/W+Q366qM+fs47uTvGAzBd390SRPT/I3ST6W5DPd/co5x2cN887tTWx/6HNgFs/IsOfJrIY8n3aceebWgPnyjMw3B+b9nW7F6+PMr4sDXg+vTHLfqjqvqs5I8rAkt9tknyeNFb2XPyM78H18C97DB89dITZJVZ2V5CVJntrd12ymtruPdPf5SW6b5MKquvsmxn14kqu6+/LNjHkT39Td90ry0CRPqqr7baJ2Tyb/hfOb3X3PJJ9N8pObbaCqTknyiCQv2mTdzZM8MpMn4ZcmObOqvnez47O2IXN7VkOeA7PYoufJrIY8n3aUeefWPPNl4ByY93c66PVxs6+L874edve7k/xyklcmuSzJFUmOzNrnyWQV7+U7+X186Hv4VszdHR9iq2pvJpP+97r7pfNuZ7r7/rVJLtpE2TcmeURVfSjJHyR5YFU9f5PjfnT671VJXpbkwk2UfyTJR477i/PFmTwZNuuhSd7S3X+/ybpvSfLB7j7Y3YeSvDTJN8wxPiewVXN7VnM+B2Yx+Hkyq4HPpx1jK+bWJufL3HNgwO906OvjZl8X53497O5ndfcF3X2/JJ/KZE3ojrLC9/Kd/D4++D186Nzd0SG2qiqTdSTv7u7/Mkf9vqo6d3r99CQPTvKeWeu7+6e6+7bdfYdMduO/prtn/iumqs6sqrOPXU/ykEx2z886/seT/G1V3XV614OSvGvW+uM8LptcSjD1N0nuU1VnTH8XD8pkLRMDDZ3bmxhn0HNgFkOfJ7Ma+nzaKYbMrXnny7xzYMjvdAteHzf7ujj362FV3Wr67+0zWVP4+5sYd/RW+V6+w9/HB7+HD527ezbz4O2mql6Q5P5JbllVH0nyM939rE1s4huTfF+Sd0zXwiTJv+3uP5mx/jZJnltVuzP5g+CF3b2ww/+cwK2TvGwyd7Inye9392Wb3MaTk/ze9L8SPpDkCZspnj7pHpzkX2xy3HT3G6vqxUnekuRwkrdm/KdN3NAWzNtZDJ3bs1r1c2ArbcXzaTQGzMMhc2vZ82Xo73Su18d5XhcHvh6+pKrOS3IoyZMGfLBnJXb4e/lo38e36D180Nx12lkAAEZnRy8nAABgnIRYAABGR4gFAGB0hFgAAEZHiAUAYHSE2G2iqr6jqrqq7rbqXhiP6Zx5/nG391TVwara1OFhqupLp4dKSVWdX1UPm6Hm/icaZ3rMwN+rqndU1ZVV9RdVdVZVnVtVPzLDdmd6HCe/qjpSVVdM59HLjx3Lc53H/2xV/avp9Z+vqm/Z4PGPqKpNn92InaOqbltV/6Oq3ldV76+qX50eympV/XxHVX31cbc3nOcnMyF2+3hckr+Y/guz+mySu08P0J1MjvX30c1soKr2dPffdfdjpnedn8k5rOf1lCR/391f2913T/JDmRwD8Nwks4TTWR/Hye9z3X3+dB59MsmTZi3s7p/u7v+9wWP+uLt/aWiTnJymB/B/aZI/6u67JPnKJGcl+Y8rbOs7knwhxM4yz09mQuw2UJPzPX9TJm/23z29b1dV/UZVvaeqXlVVf1JVj5l+7YKqel1VXV5Vf1pVt1lh+6zenyT5tun1LzrrSlVdWFVvqKq3VtVfHTurS1X9QFX9cVW9Jsmrq+oO071dpyT5+SSPne4Be+xa21jHbXJckO7u93b3jUl+Kcmdptt92nTv7Kur6i3TvbaPnJbc9HFftMe3qn69qn5gev2XqupdVfX2qnr6gJ8h298bknxZklTVnarqsulr4J+f6H+wquo5x71mPmz6Wnp5VT3z2HyaPg9+fXr9DlX1mulcenVNziB0bDvPnM79DxzbJjvCA5Pc0N2/kyTdfSTJjyf5wZqcaevp09fNt1fVk5Okqu49nStvq6o3VdXZx8+z6WNeUVX3n16/rqr+a1W9czrv9k3vf2JVvXm6nZfU5H+4viHJI5I8bfr6eKebzPMHTV+n31FVz66qU6f3f6iqfu6419qT5n98hdjt4ZFJLuvuv07yiaq6IJPTr90hk7+4vi/J1ydfOD/0ryV5THdfkOTZWe1fhazeHyT57qo6LcnXJXnjcV97T5L7dvc9k/x0kv903Nfulck8+uZjd3T356eP+8PpHrA/3GAbJ/LsJP9mGnx/oaruMr3/J5O8f7rdS5LckOQ7u/teSR6Q5D9P93zc9HEnVJOzvHxnkq/p7q9L8gsb9MVI1eRMSg9K8sfTuy5N8uTpa+C/SvIb69SeluS3kjx0+vh9azz015I8dzqXfi/JM4/72m0y2dHw8Ez+yGJn+Joklx9/R3dfk8npVv95Ju/R5x+bM9OdAH+Y5CndfY8k35LkcxuMcWaSA939NUlel+Rnpve/tLvvPd3Ou5P8UHf/VSbPgUumr4/vP7aR6Tx/TpLHdvfXZnL2rx8+bpyrp6+1v5nJc+akMOrTzp5EHpfkV6fX/2B6e0+SF3X30SQfr6rXTr9+1yR3T/Kqyft9dif52HLbZTvp7rdX1R0ymTc3Pc3izTI5neJdknSSvcd97VXd/ckZhlhvGyfq54qqumMm5wD/liRvrqqvzz9+Ma8k/6mq7pfkaCZ72W49Qz/HfCaTIPys6Z61sZ7ulrWdXpPTiH5ZJm/kr6rJ/1x9Q5IXTV8Dk+TUdbZxtyQf6O4PTm+/IMnFJ3jc12ey8yBJfjfJrxz3tT+avha/q6o2M0c5ed0/yW909+Ek6e5PVtXXJvlYd795et81SXLcPD2Ro5kE3yR5fibLF5LJMrFfyGR51VlJ/nSDfu6a5IPTnWFJ8txMlt88Y3r72HYvzz/M89ETYlesqm6RyX9ZfG1VdSahtJO8bK2SJO/s7q9fUouMwx8neXomL6znHXf/f0jy2u7+zmnQ/bPjvvbZGbe93jZOqLuvy+RF86VVdTSTNbYvucnDvieTvWIXdPehqvpQktNOsLnD+eL/NTptOsbhqrowkz10j0nyo5k8lzh5fK67z6+qMzJ5E39SJnubPt3d5y+xjxuPu75uIuGk8q5MXlu+oKrOSXL7JB/axHZO+Bq2hp7++5wk39Hdb5sun7r/JsY7kWNz+EhOouxnOcHqPSbJ73b3l3f3Hbr7dkk+mMmHGB5dk7Wxt84/TOD3Jtk33bOVqtpbVV+zisbZVp6d5Oe6+x03uf9m+Yf1qT8w47auTXL2vNuoqm+sqptPr5+SyZKYD6+x3aumAfYBSb58jfE/nOSrq+rUmnw6/UHTbZ+V5Gbd/SeZrFO7x4zfHyPT3dcn+bEk/zLJ9Uk+WFX/NJl8+Kaq1vvdvzfJHad/gCXJY9d43F9l+pmETP7A+vOhfTN6r05yRlU9PvnCspb/nEnA/NMk/6Kq9ky/dotM5tptqure0/vOnn79Q0nOn76f3y7JhceNsSv/EJT/WSYf8E4mr4Efmy4h/J7jHn/T18dj3pvkDlV15+nt78tkecJJTYhdvcflH+91fUmSL0nykUz+Enx+krck+cx0zeJjkvxyVb0tyRWZ/NcaO1h3f6S7n3mCL/1Kkl+sqrdm9r++X5tJaLyiqh47xzbulOR1VfWOJG9NciDJS7r7E0n+cvpBiKdlsu5w//Rxj89k7W1u+rju/tskL0xy5fTft07HOTvJK6rq7Zm88P/EjN8fI9Tdb03y9kxeM78nyQ9NXwPfmcnnCtaq+1wmR7u4rKouzyQEfOYED31ykidM59P3ZXKUDXaw7u5M1t3/06p6X5K/zmQJ079N8tuZrI19+3Qe/rPp+/Njk/za9L5XZbLX9S8z2Tn1rkzWWr/luGE+m+TCqroyk/9J+vnp/f8+k883/GWmr41Tf5DkkukHuO50XK83JHlCJsts3pHJMoX/tlU/i+2qJr8jtqOqOqu7r5t+gOVNSb6xuz++6r4AxuS419JK8v8leV93/9dV9wVVdV13n7XqPsbqpFkXcZJ6xfS/T09J8h8EWIC5PLGqvj+T19K3ZnK0AmDk7IkFAGB0rIkFAGB0hFgAAEZHiAUAYHSEWAAARkeIBQBgdP5/Ig6uo3MQxcoAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "