From 6ac4d0a250212f5e37438a742ba59196cf7886dc Mon Sep 17 00:00:00 2001 From: NoidFrancis Date: Thu, 11 Sep 2025 00:37:39 +0200 Subject: [PATCH] Lab hypothesis testing done --- lab-hypothesis-testing.ipynb | 551 +++++++++++++++++++++++++++++++++-- 1 file changed, 533 insertions(+), 18 deletions(-) diff --git a/lab-hypothesis-testing.ipynb b/lab-hypothesis-testing.ipynb index 0cc26d5..f1bcd80 100644 --- a/lab-hypothesis-testing.ipynb +++ b/lab-hypothesis-testing.ipynb @@ -38,7 +38,7 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 8, "metadata": {}, "outputs": [], "source": [ @@ -46,12 +46,12 @@ "import pandas as pd\n", "import scipy.stats as st\n", "import numpy as np\n", - "\n" + "from scipy import stats\n" ] }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -278,7 +278,7 @@ "[800 rows x 11 columns]" ] }, - "execution_count": 3, + "execution_count": 9, "metadata": {}, "output_type": "execute_result" } @@ -297,11 +297,46 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 10, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "n (Dragon, Grass): 50, 95\n", + "Mean HP (Dragon, Grass): 82.90, 66.05\n", + "Welch t = 4.098\n", + "One-tailed p (Dragon > Grass) = 5.091e-05\n", + "Decision @ 5%: Reject H0 (evidence Dragons have higher HP)\n" + ] + } + ], "source": [ - "#code here" + "# Groups: Type 1 OR Type 2 matches\n", + "is_dragon = (df[\"Type 1\"].eq(\"Dragon\")) | (df[\"Type 2\"].eq(\"Dragon\"))\n", + "is_grass = (df[\"Type 1\"].eq(\"Grass\")) | (df[\"Type 2\"].eq(\"Grass\"))\n", + "\n", + "hp_dragon = df.loc[is_dragon, \"HP\"].dropna()\n", + "hp_grass = df.loc[is_grass, \"HP\"].dropna()\n", + "\n", + "# Welch t-test (two-tailed first)\n", + "t_stat, p_two = stats.ttest_ind(hp_dragon, hp_grass, equal_var=False)\n", + "\n", + "# Convert to one-tailed for alternative: mean(Dragon) > mean(Grass)\n", + "diff = hp_dragon.mean() - hp_grass.mean()\n", + "p_one = p_two / 2 if diff > 0 else 1 - p_two / 2\n", + "\n", + "print(f\"n (Dragon, Grass): {len(hp_dragon)}, {len(hp_grass)}\")\n", + "print(f\"Mean HP (Dragon, Grass): {hp_dragon.mean():.2f}, {hp_grass.mean():.2f}\")\n", + "print(f\"Welch t = {t_stat:.3f}\")\n", + "print(f\"One-tailed p (Dragon > Grass) = {p_one:.4g}\")\n", + "\n", + "alpha = 0.05\n", + "print(\"Decision @ 5%:\", \n", + " \"Reject H0 (evidence Dragons have higher HP)\" if (p_one < alpha and diff > 0) \n", + " else \"Fail to reject H0 (not enough evidence)\")\n", + "\n" ] }, { @@ -313,11 +348,417 @@ }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 11, "metadata": {}, - "outputs": [], + "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", + "
StatMean LegendaryMean Non-LegendaryWelch tp (two-sided)Hedges g (Leg − Non)n_Legn_Non
0Sp. Atk122.18461568.45442213.4174501.551461e-211.83468565735
1Speed100.18461565.45578211.4750441.049016e-181.26246365735
2Attack116.67692375.66938810.4381342.520372e-161.34418165735
3Sp. Def105.93846268.89251710.0156972.294933e-151.42697665735
4HP92.73846267.1823138.9813701.002691e-131.03892265735
5Defense99.66153871.5591847.6370784.826998e-110.92840165735
\n", + "
" + ], + "text/plain": [ + " Stat Mean Legendary Mean Non-Legendary Welch t p (two-sided) \\\n", + "0 Sp. Atk 122.184615 68.454422 13.417450 1.551461e-21 \n", + "1 Speed 100.184615 65.455782 11.475044 1.049016e-18 \n", + "2 Attack 116.676923 75.669388 10.438134 2.520372e-16 \n", + "3 Sp. Def 105.938462 68.892517 10.015697 2.294933e-15 \n", + "4 HP 92.738462 67.182313 8.981370 1.002691e-13 \n", + "5 Defense 99.661538 71.559184 7.637078 4.826998e-11 \n", + "\n", + " Hedges g (Leg − Non) n_Leg n_Non \n", + "0 1.834685 65 735 \n", + "1 1.262463 65 735 \n", + "2 1.344181 65 735 \n", + "3 1.426976 65 735 \n", + "4 1.038922 65 735 \n", + "5 0.928401 65 735 " + ] + }, + "execution_count": 11, + "metadata": {}, + "output_type": "execute_result" + } + ], "source": [ - "#code here" + "from math import sqrt\n", + "\n", + "stats_cols = [\"HP\", \"Attack\", \"Defense\", \"Sp. Atk\", \"Sp. Def\", \"Speed\"]\n", + "\n", + "leg = df[df[\"Legendary\"] == True]\n", + "non = df[df[\"Legendary\"] == False]\n", + "\n", + "def hedges_g(a, b):\n", + " # unbiased standardized mean difference (like Cohen's d with small-sample correction)\n", + " na, nb = len(a), len(b)\n", + " sa2, sb2 = np.var(a, ddof=1), np.var(b, ddof=1)\n", + " # pooled SD with unequal n (still used for effect size, not for test)\n", + " s_pooled = sqrt(((na-1)*sa2 + (nb-1)*sb2) / (na+nb-2))\n", + " d = (np.mean(a) - np.mean(b)) / s_pooled\n", + " J = 1 - (3/(4*(na+nb)-9)) # small sample correction\n", + " return J * d\n", + "\n", + "rows = []\n", + "for col in stats_cols:\n", + " a = leg[col].dropna()\n", + " b = non[col].dropna()\n", + " t, p_two = stats.ttest_ind(a, b, equal_var=False) # Welch\n", + " g = hedges_g(a, b)\n", + " rows.append({\n", + " \"Stat\": col,\n", + " \"Mean Legendary\": a.mean(),\n", + " \"Mean Non-Legendary\": b.mean(),\n", + " \"Welch t\": t,\n", + " \"p (two-sided)\": p_two,\n", + " \"Hedges g (Leg − Non)\": g,\n", + " \"n_Leg\": len(a),\n", + " \"n_Non\": len(b),\n", + " })\n", + "\n", + "res = pd.DataFrame(rows).sort_values(\"p (two-sided)\").reset_index(drop=True)\n", + "res\n" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Benjamini–Hochberg critical p (max rejected) at q=0.05: 4.826998494919331e-11\n" + ] + }, + { + "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", + "
StatMean LegendaryMean Non-LegendaryWelch tp (two-sided)Hedges g (Leg − Non)n_Legn_NonBH q=0.05 significant?
0Sp. Atk122.18461568.45442213.4174501.551461e-211.83468565735True
1Speed100.18461565.45578211.4750441.049016e-181.26246365735True
2Attack116.67692375.66938810.4381342.520372e-161.34418165735True
3Sp. Def105.93846268.89251710.0156972.294933e-151.42697665735True
4HP92.73846267.1823138.9813701.002691e-131.03892265735True
5Defense99.66153871.5591847.6370784.826998e-110.92840165735True
\n", + "
" + ], + "text/plain": [ + " Stat Mean Legendary Mean Non-Legendary Welch t p (two-sided) \\\n", + "0 Sp. Atk 122.184615 68.454422 13.417450 1.551461e-21 \n", + "1 Speed 100.184615 65.455782 11.475044 1.049016e-18 \n", + "2 Attack 116.676923 75.669388 10.438134 2.520372e-16 \n", + "3 Sp. Def 105.938462 68.892517 10.015697 2.294933e-15 \n", + "4 HP 92.738462 67.182313 8.981370 1.002691e-13 \n", + "5 Defense 99.661538 71.559184 7.637078 4.826998e-11 \n", + "\n", + " Hedges g (Leg − Non) n_Leg n_Non BH q=0.05 significant? \n", + "0 1.834685 65 735 True \n", + "1 1.262463 65 735 True \n", + "2 1.344181 65 735 True \n", + "3 1.426976 65 735 True \n", + "4 1.038922 65 735 True \n", + "5 0.928401 65 735 True " + ] + }, + "execution_count": 12, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "m = len(res)\n", + "p = res[\"p (two-sided)\"].values\n", + "order = np.argsort(p)\n", + "ranked_p = p[order]\n", + "bh_thresholds = 0.05 * (np.arange(1, m+1) / m)\n", + "bh_significant_mask = ranked_p <= bh_thresholds\n", + "# largest index that satisfies the condition\n", + "k = np.where(bh_significant_mask)[0].max()+1 if bh_significant_mask.any() else 0\n", + "p_bh_crit = ranked_p[k-1] if k>0 else np.nan\n", + "\n", + "res[\"BH q=0.05 significant?\"] = res[\"p (two-sided)\"] <= (0.05 * (res.index+1) / m)\n", + "print(\"Benjamini–Hochberg critical p (max rejected) at q=0.05:\", p_bh_crit)\n", + "res\n" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Sp. Atk | mean Leg=122.18, mean Non=68.45 | t=13.42, p=1.55e-21 | g=1.83 (higher) | SIGNIFICANT\n", + "Speed | mean Leg=100.18, mean Non=65.46 | t=11.48, p=1.05e-18 | g=1.26 (higher) | SIGNIFICANT\n", + "Attack | mean Leg=116.68, mean Non=75.67 | t=10.44, p=2.52e-16 | g=1.34 (higher) | SIGNIFICANT\n", + "Sp. Def | mean Leg=105.94, mean Non=68.89 | t=10.02, p=2.29e-15 | g=1.43 (higher) | SIGNIFICANT\n", + "HP | mean Leg=92.74, mean Non=67.18 | t=8.98, p=1e-13 | g=1.04 (higher) | SIGNIFICANT\n", + "Defense | mean Leg=99.66, mean Non=71.56 | t=7.64, p=4.83e-11 | g=0.93 (higher) | SIGNIFICANT\n" + ] + } + ], + "source": [ + "# Neat printout of conclusions\n", + "alpha = 0.05\n", + "for _, r in res.iterrows():\n", + " direction = \"higher\" if r[\"Hedges g (Leg − Non)\"] > 0 else \"lower\"\n", + " sig = \"SIGNIFICANT\" if r[\"BH q=0.05 significant?\"] else \"ns\"\n", + " print(f\"{r['Stat']:<8} | mean Leg={r['Mean Legendary']:.2f}, \"\n", + " f\"mean Non={r['Mean Non-Legendary']:.2f} | t={r['Welch t']:.2f}, \"\n", + " f\"p={r['p (two-sided)']:.3g} | g={r['Hedges g (Leg − Non)']:.2f} ({direction}) | {sig}\")\n" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "C:\\Users\\noidf\\AppData\\Local\\Temp\\ipykernel_22984\\355455809.py:13: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(data=subset, x=\"Group\", y=\"HP\", palette=\"Set2\")\n", + "C:\\Users\\noidf\\AppData\\Local\\Temp\\ipykernel_22984\\355455809.py:18: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.violinplot(data=subset, x=\"Group\", y=\"HP\", palette=\"Set2\", inner=\"box\")\n" + ] + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAHqCAYAAADVi/1VAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjUsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvWftoOwAAAAlwSFlzAAAPYQAAD2EBqD+naQAAtZdJREFUeJzs3QWYVGXbB/D/5HYHuaR0d8jHq4AiYID1qtgIomAAoqKCLQYKSogCoqiYryCogAgiqHRLdyy53TnzXfcDZ51dttndM/H/Xdcw7MzszLNnTjznPvdzPwa73W4HERERERERERFRFTJW5YcREREREREREREJBqWIiIiIiIiIiKjKMShFRERERERERERVjkEpIiIiIiIiIiKqcgxKERERERERERFRlWNQioiIiIiIiIiIqhyDUkREREREREREVOUYlCIiIiIiIiIioirHoBQREREREREREVU5BqXIoxkMBrz00kt6N8OpVdUyWrVqlfosuddcddVVaNmyJarC0aNH1ed/+umn0Ev//v0xdOhQ3T6fXFPXrl3x9NNP690MIqpAciySY5Icm8rq/vvvR7169ZyivyOfKZ9NzrGMpF8lt4J9r++//75KPr+wdbMqbdiwAVarFceOHdOtDeR6li5dCn9/f5w/f17vprgtBqWoUjtTjrfIyEhcffXVWLJkCVzd7t27VSeiPJ1FR8UFXbQgyaRJky7pPGg3i8WCBg0a4N5778Xhw4dL/DzpCGi/azQaERwcjFatWmHYsGFYv349Ksr8+fMxZcoUOCNnbdtff/2FX3/9Fc8880yR37eXlxeqVaum1ps33niDB8cKlpSUhNdffx0dO3ZEUFCQWt5169bFf//7X/z8889wVrLOTJ8+HWfOnNG7KURUhBtvvBG+vr5ITk4u8jWDBw9WJ8yxsbFwdhJccDw+BQYGok2bNnj33XeRmZlZIZ8xY8aMy75QVFLQRf4OOdl0JMdYx78tNDQUnTp1wieffAKbzVam/q+3tzdq1qyJvn374oMPPij2+y+LU6dOqX7otm3b4GycuW3PP/887rzzTnVsL+z7lr6xrMtNmjTBPffcg+XLl+vaXne0Zs0a3H777ahVq5ba30l/q0uXLnjllVdw9uxZOKPrrrsOV1xxBSZOnKh3U9yWWe8GkHuTHUz9+vVht9vVjkYO1pINsnjxYlx//fVw5aDUyy+/rA5kelzxefzxx1UHKTs7G1u2bMHHH3+sTpp37typOj/Fadu2LcaMGaP+L52jPXv24LvvvsOsWbMwatQovPfee/len56eDrPZXObAzz///IMnn3yy1L/Ts2dP9VlygKpMRbVNOijy+RLo08M777yD3r17q4NeUd93bm6uCkT9/fffePHFF9V39e2336JXr166tNmdHDx4UJ00yNXTQYMGqUCvnKicOHECv/zyi9pfzZs3T3VSnc1NN92kOtFyAif7XCJyPhJwkr7PggUL1P6loLS0NPz444/q5CcsLEzta+644w4VHK8I5TmWl0TaNnv2bPX/hIQE/O9//8NTTz2FjRs34uuvv77s95d9Wnh4uAocVbXatWvnnYDKcVf2/0OGDMH+/fvx5ptvlrr/K/00uWAgwTHpd8hxe9GiRWjdunXea1944QU8++yzZQ78SD9U+qDSrystufhV2Yprm/Q1SwrsVRYJkv3222+qD1Xc952amqr6BD/88AO++OILFUCRe736h+5kwoQJePXVV9UFddmu5T4jIwObN29WAe3PPvsMhw4dgjN6+OGH1f5N1u2AgAC9m+N2GJSiStWvXz+VdaCRA7pkenz11VcuHZTS2//93//h1ltvVf9/4IEH0LhxYxW4kJ35uHHjiv1duTJx991353vsrbfewl133YXJkyejUaNGeOSRR/Kek6t8lUkORhKIkqtTlf1ZxdGuaOrh3LlzKqg4c+bMEr9vzfbt23HttdfilltuUUHSGjVqFPn+0sHy8/Or8Ha7i5ycHBWIksD5H3/8gSuvvDLf8xIAlI68BAWLo9dylm1H1g85aZLOEofKEDlnppScyMiFkcKCUhKQkn2IBK+EyWRSt4pSGcc3CXI59iceffRRlfHwzTffqOBLSRfJnJlkbzj+bXJCKtkz06ZNUyfVJQUoCvZ/pW+2cuVK1feVdUEuCPr4+OQtx4oOGBYW9JRMvcq+8FcSPQM7c+fORZ06ddSQ95K+byHBR+lbS3BUAmzSVy6KBNqysrJ07cc6O9kvyLYjQb7PP//8knVRzkHkVhxJcpDzBm3bqUrS337sscfUhfwHH3ywyj/f3XH4HlUpGS4mO5KCB1/piEn2TlRUlLryJgd+GbYmOx/tCl/Tpk3VTf6viYuLUyfj3bt3zzth1FKxZTibZD7ISaJ0jOSqlfZ+xdm6davqTEjmgbyPZK+sW7cu73nJ9rrtttvU/2U4opbyq9VCSkxMxN69e9V9VdEyZY4cOVKu35fvRA4QkqIuw5ccl1PBOhSSXSVX++QALd+VDMu85pprVMaWkOwxCbBIxom2bLRsMi2NXq6gypVBCZBJJ0mGTRVWU0ojV1DkO5Z2ypXHgsGbompvFHzP4tpWVE0p6URKUEjWI1l/JStFOpOF1YOQK2uy/snrpIMjAUPpCJZE2iSBkT59+qC0ZJiEDEOUq9PSSS7YFglUSaAxJCQEPXr0UM/t2LEj78qUdJyqV6+uDqyFDRWRZSYdanldw4YN8dFHHxVa90LaLZ0MeY2sD7I8n3vuuUuGb8jj0hn/888/0blzZ/W+0g4JpBRHrjLLeinLsiBZb+R95MqVZurUqWjRooVar+Rvl79BTgKLIx0MyZ4bP378JQEpjQQAZb9QcJ2TIJaciMl2IFdahaxf8pjsx2SdlawH2WcUXD/lb5MgkgSC5e+Q18l35ThcQK6wy98u7y3LV/Z3sg4WfC/ZBuVznXG4BBFdOM7efPPNWLFihboQUZDspyRoJQGL4o5rcoIs+zjZH0jfZsSIEeo4UJKCx/LLPW4VFSDX6hUVV96gNMcNeWzXrl1qH6sdrx1rIUk2RVVmVMgxRYIZ0l8t79B56avJcUb21ZJ5oyns2CrHATkeyPcifVE5nsgy0o7Pkj0t5PvSlo/Wf9FKQ0jfSbLQpe3a7xasKaWRPrS8RvoF0t+R9VAyhR3Jd1JY1prje5bUtsJqSpV0DqCR9xk5ciQWLlyo/j55rWwLUu+nNOT35Dso7YUbCQrLkMvmzZurfpZjv15ry5dffpm3PWrtkLZLn1WO6bLdd+jQodDho3I+I0EvyQbUtv3o6OhC67+VdG7iuM+QchCjR49GRESE+i7loltJ66y0WX63sFpbElCVAFJ8fLz6+cCBAypAI+uK9F2kfyJZnSWd90iWlPytc+bMKTQ4Kvufgn+31ndctmyZ6s/J8pT+qBZklO9T+l+y/OV7+vDDDy95302bNqnzQfls7TyiYFBJzkvke5LvQZaxlDZ5//33871GPkcyHOUCAlU8ZkpRpZIdVExMjDqwSCdMThhTUlLyXY2Q52RH/Pvvv6tMKkn1lZ3P2LFj1c5ZouayE5EsIDlhlPHg2hAz6YzJZ8iO2PGKohxcJQVeOhBvv/22OlBItoN0hIob3iIdIAlAyA5JCgfLFR3Z+cnBVjpGcgVQDvByEJEDlRzAmzVrpn5Xu5fUfDkQy86yNCnn0lZZRgVpO//S0DpmcgAsLznIyYFLDhYS0JCDbGGGDx+uDq5yMJYDgAQ0JNAggZr27dur70e+k5MnT+Zd8ShYr0E6o3JAkmCCdEKLu3Iny0GGfMqVFakDIMPVJJNLfqesVypK0zZHkuYtnQAJnsiBUjoQsg7LeihBuIIdK2mjHOwkBVyel2ENchAr7uqakFRy+e4caxyUhmTHyDYjWTwSTHQkQRAJdkjtKa1jJ51cCdbK+imdCVnfZein3EvnRuuoSedHth8JgEjQRNZR2W6kg1PQQw89pLZNaYt0KqU2mfz9sj7ItuBITn60Nt93332qPodsI9IRKGp9k21Q1ktJo5dt0XFdkQ6mrD/SGdKGBci2KZ/xxBNPqKtpEoiTNkmArigypEYUvEpaGhJ8kuUinS3pWAsZuiLfqbRLOmtyciYdJdmPyLYlJwhC1ilZVrIMJVAnQTbpPMm6I0EmIR0/+X7k6pysb7Ifle/x+PHj+dY/WYZCOqPt2rUr899BRJVPsqBkfynHMTmGOl5gk36PHOOKywCQfYbsk+UChhwH9+3bp/Ytss+Rbb88WSjlPW5dTn+kNMcNuegi+z05RsuxW0imvUZOykVpa3vKBbXC+lplqX8lx0/pa0qgqLxkWKb0HeW4XdTEJrLPlxNxOQGWY6+ccMvxU75jrb8pj8txR2qCSr9VSCBEI30z6b/IcUiObY7LrjDSh5A+gNQolOOMLH9Zz+RCR1myUkrTNkelOQdwJP1N6Q/IsVcCCNIXl+OkHBOLW+fkveQ10k8tC/m+ZbuUYKJ89oABA/JdtNS2ZQl4aMdkCWbI3yTbu2RPScBD+mQ//fRTvt+X/o/8vqwTcr4i5xmOz5fl3MSRbDdyUU7Oe2T7kO9S2iiZSsXtB+S9pT2y7B3JY3JhTt5T/h4J8Mh2I58jfUlZtvK3SXBcAkuFkWGvcpNtv7h+d2FkPyffgWQryjYjQUsh+z7pO8qylmQH6cvJeiFZa3J+KGRdlrZLP02GyMq2K8tE1iGN9Knk/WWfou33ZF8k25v0JR1JX0v6nlQJ7ESVYO7cuXIWfMnNy8vL/umnn+Z77cKFC9Vzr732Wr7Hb731VrvBYLAfPHgw77Fx48bZjUajffXq1fbvvvtO/d6UKVPy/d59992nHn/sscfyHrPZbPYBAwbYrVar/fz583mPy+tefPHFvJ8HDhyoXnPo0KG8x06dOmUPCAiw9+zZM+8x7bN///33Iv92uS/Jf/7zn0KXk+PtnXfeyXu9fJ489sknn6i/Q9r2888/2+vVq6eW1caNG4v9vLp166rlUJTJkyer9//xxx+LXEZBQUH2ESNGFPs58hnyWQVp7W/QoIE9LS2t0Occl6m2fN599928xzIzM+1t27a1R0ZG2rOysvIt8yNHjpT4nkW1TX634PemfU5sbGzeY9u3b1fr4L333pv3mCwf+d0HH3ww33sOGjTIHhYWZi9Jjx497B06dChyecn6VpQ2bdrYQ0JCLmnLnXfeeclrCy5z8dVXX6nXyzalueGGG+y+vr726OjovMcOHDhgN5vN6rWabdu2qZ8feuihfO/51FNPqcdXrlyZ95gs84Kfc+7cObVPGDNmjL04y5YtU7+7ePHifI/3799frUuam266yd6iRQt7WbVr184eHBx8yeMpKSlqO9NuiYmJec9p65x8dzk5OSUu57Vr16rXz5s3L993V9z2GB8ff8k+oDiy73rkkUdK9Voiqnqyr6hRo4a9W7du+R6fOXOm2tZlX6cpeFyT/aVs49dee609Nzc373XTpk3L6xc49oMKHucKHssv97gln+Hn55e3f5S+2htvvKH6Iq1bt77kc8pz3JD9ufQDCiN/X2HH8qKOo8Xd5O9wJJ/ZtGnTvL9tz5499scff1y9Vo6PxdG+t+L6Y9KPkuNOUctI64s59lcLkvcvqq+p9Z1kvSrsOcdlqi2fWrVq2ZOSkvIe//bbb9Xj77//ft5jsrzley/pPYtrW8F1syznAPI62QYcH5M+mTw+depUe3F+++23QvsRWvuL6zssWLDgkmUhP0tfcNeuXZe8vmAfQPqqLVu2tPfq1Svvsc2bN6v3ePLJJ/O99v777y/3uYm27vXp00ed92hGjRplN5lM9oSEBHtxZL9UsC+6YcOGfH2XrVu3ltgvLYycVxR2zibtdOxnyS07O/uSvuPSpUsvec/C+lp9+/bN1y/UvrvitscnnnjCHhgYeElfrjCyj5P3O3v2bImvpbLh8D2qVDIjlESg5SapyjLcTaLkjhFqKSIsVyIkw8GRXD2T/b7jbH1ylVCi4pJlIdHw//znP5f8nsbxKqSWZisRfsl+KYxkg8iVq4EDB6rMGI1ki0iWhVwhkUyGksiVD2l3aQtzypUVbRk53hxTuwuSDCGJ+kvqvlxVkQwNueroWL+gPLSrF8XNDiNXGeSqphSyLC/5/kp75U2ufsjVEY1kysjPcvVDUtMry+nTp9UVQvkeZfiYRq5cShaLrLeFZZE5kitbcrWypPVGXiNXoMr7nRX2fRVsi3Bc5pJFJFeNtdoK2vBL2Q5kG5HtwLEeiBRgdxy+JrRlIGnijrRC+gVnrJPMOu2qqZB1WK54lTRzpKRny1VIx6t8kkEn24nMjOe4bkoWnGQNlIV8P4VduZOr89JG7VZYtpVctStY98VxOcsQPfl+ZflJ+7TlrLVXroBKKnxh5H1kfZfhEKXJnJR1qLBMACJyDrKvkMyVtWvX5svwkaF7ksmiZf8URvbL0oeR4fMyTM5xHyQZFOWdIbS8xy0hfQ9t/yj7OMkA6tat2yVZspdz3CiKLL+yzIAsmTuF9bUki6IwUoZB+9sk+0eypKW/JRm+l6uo47ZGy8SSYULlLQou2VWFDXsvitQ5cyzeLFls0v8trK9TkcpyDiAke0uGfTr2yWT9L6kfoZUpKE9fq6i+sZyDSL+mIMc+gBy7JUNftivH47821E/OZRxJ9tHlnptIhprjEEX5bHmfwobmOZL+lPSrHYfFSr9L1iUpGyC0TCjJZivLMF+tjQX7WrJsHPtZcitYhkAyOSU7q7jlrI3Mke9E1gVtKKG2LUkml/THCiOvkX1ZaWZa1NYf9rUqHoNSVKlkSIocQOQmaazS2ZAduBYgErKTlJPfgjMZaMPhHHeicoImHQKpnSQHBxkiV9jYcOmwOe68hRQDF0V1YmS8texgtbTQgm2RjkHB8fUVQcZ7a8vI8VZUbRvHzpWkDsvwJAkQVcSsYDK0UhQ3q4QMh5T6OzL2X75fCRSW1BkoSA4wpSXrRsHi0SV9lxVBW++KWh/kgKQN19JIAc3CDl6lCSiUpt5ZUd9ZYd9XYctYhohIKrKc/MjBXA7+2uu0A7gE+2SYYmGzABZ8TJaRbGsFH5d0bjnIF+wAFVw+2jIqaflIYFLS86WDrg21kMC2dDAcg1Iy7EA6PLJeytBFSd/WhjsUR5aftu47ks6iduJS1NCHwpazLD/ZRrX6GBJQk2Utqe2ONRdkiIM8Juuz1C+QlHnZnjXyu5JKLp1y+XwZOizbn9SZKmodYpFzIuemFTLXat1JIF2mSJdgVXGFzYs6Jkm/SPo7JZ1wFuVyjltST0bbR65evVr1kWSfW7D/dTnHjYoi+9jC+lpFTRKiXTCUYKCc+Mt+V05sZX9+uYo6bmvkuCZ9QLmIK/t+WTdkCFVZAlRSs7MsRc3lmOlIjiXyHVVmP6us5wCX04+4nL5WUX3jovqysp7IBT/ZPuSiphz/ZaiZ4/Ff2w4KvkfB7aI85ybl3aZliKG0SbsAKMtKam5qtay0v1kCyjLMV7YFCRZJAkJJ9aS0ZVewryV9Nm0fUnDYYEnLWfY1sg1rNV9lOWu107T2SJBK+o8y7FnaK8E1OXd0HLYrfT3ph8nfKSUX5MJ/UXXKtPWHfa2Kx6AUVSnZ2Um2lGShFJUdUBKJzmuZHuV9D1enda5kWcr/K2rWFgk2icICEo7jziUIJVcNpSPxzjvvqOy1glezilPRs2YUdXAoaba0ilbUCUVJnSCpg1CWGmIaCcrIGP3Cvq/ClrF8d1J3Sa6MS1BHrr5pB97LmaK5LEVDy9tJlE65BKK19Uw66DLxgRR8d+ygSe0Bqd8gBWJlenK5l7oKxZH3keCQ1EVwJJ0U7cSlqBl1ClvOcqVT6nPI8pZ2ynKWDpd8z47LWYJMckVSAu1StFU6eVLvQptiXUhWhHzHUm9F2iB1LeTvlLpfBcnfUBEnTERUeaQmiexzZBZiIfeyD9SCVVXtcvbL8rvaPlKyMbTJHkrD2U/qtAuGkr0mASKps1URJAgpJ8zF9bPkuCJBPgmIyQVHuVghgSrJ0i5tv6YyZidzhr7W5fSzRHn6WkX1jQtbxhJglhpHcryWSQkkE0yO/5LVVN6Lj1W1jKRPL9ux9FuE1BqVOlyOF//Eu+++q9ZJCQBpxdrlPEDW7aLIPs9xWWrk/EXbhxSWdVbUcpa+k2ybcoFY6gxL0oMs51GjRqnntb6WrLNSB1eyUyUhQvp5EnSS/bAWIJNtW7KzFi1alFffTAJUMqqjIG39YV+r4jEoRVVOio0LbWcgxZ0l06dgWqykTmvPa2QnKNkFkpIsxXzlKlJh0XnZGRXM3pETO1GwOLVGIuxSgFhOaguStkhATTIfXKEzVR7yfUjKvfyN2hWqosiVRbmyIMX+JGtNDvaOhbYrcvnIulEwI6ngd6ldBSo4A1FhV1xL2zZtvStqfZADUsEMrvKSg3V5Zk6UA610CApLay7sQCqzPkmhR7liJMXDpYNb8Iq2HJylMyVFVQsq+JgsI9nWCgaHz549q76LshZuL44EcGS9kyt40gmRLMGCHSUh34k8LlfCpDMlwy1k3ZQgdlGkoKyQWXQqgnwv0pmRjpsMgZDlLMGxwmbI0mYWlBNTudopQxEKzj4jQxVkKIMEt6RDJ1mm8t6OpKMlj5e07RKR/iQAJduy9GkkY0qyVLQZy8p6TJLtXo4fFbm/rUxlOW64Y19LZjoWJR23pc8pJ91ywi0TZMhxTI57csJcGcum4PchAQw55jv2maWvVdhxrGBfqyxtK8s5wOXQgiJl7WtJwE22UTk/0GYyLo5cDJM+lFxAl+CHBDcKm1lZ2w4KtqdgP6ss5yYVQfpP27dvV58n/S357BtuuOGS18kFcZlFW4KnEoiTPkjBmbEdSaaX7OfkvKFgn748pKi5ZDtJIElKesiESLKciwrGSuaabEMymYz09aR0glzA1EhWofydEkiUgJe8p8wOXfD7kO9Ly36nisWgFFUpyeyQEyvZ+LWTJ9mRyE7fcVp7ITNuyIFNq2Mjvyv1fSSSLzNbyIx70onRouIFOb6fHFzlZ5mxoqiaDXJlQWoLyBAhx3Rl+Qw5IMnBSEtf1YIRhR2cJUgmB4qSUlmdiQQ25GqcDO+SOjrFXQ0r+HdJEEO+E8dUWFk+FfX3SxBTm/5V64DLz3JA0GYc0+oLyMHRsa0ys1xBpW2bBEBkFhip1eX4PcuJhKzDst5WFKnBIUGjsgyDlE6DZNFIJ1GbZaQ0V84KXimTWVkKvk4O7NJxcKwbJgfmgtlw2jIo+B7a7JiFzSJTXtLxkgCPdESkUy/rRcGglFYzQiP7GbnyJn9zUbUEhGQ0yetkVsiCUyxrynKFU5ZhwddLZmHBq8kF2ytp7HIlVtuWJGW/YDBN1nVJgy84Y5RWX62oGY6IyHloWVEyzFeu0JcmS0r2y7JPk9nGHPcvMmOuHNMqcn9bmcpy3JDjdWH9LCEnjo61b1yBBJXkOCPDkYr7zqUvVpD0R4S27y+uH1oecgLuGBiSiysyqsGxlqQcf+QYqZXf0IaqFRw+Vpa2lfYc4HLJcEYJ3khQorSkXZIFJDOxyb12DlDS8V/a7Xi8l3OKgjO2aUFJCYIU7CuU99ykIshQN/lMuVAmQ/fkop3jBVipDaUlFzgGqKSPVtJMlnLBTS4qSh28wvpkZe1nFfwd2Q/KBUlH0rcu+L4Ft6WCfTH5W+QCoeNrHPta0menilcxY36IiiAnsdrVDqlVIztQuRoj2RraTlQi0zIMTYIhssOV4Thy0i87YDnp1gIOr732muq8SbaHnJTJDkM6dBKpl5NVxyCBXKWQYUmSrSBTpUo7JLVTUk2Li27LZ0j6p+zkJRNI0kolACI7Janl4rhDkx2i1HuRnaDUfpFizBKgkWwjyXyQHWNpi51XJbmaoRVRl+wouQInBx6plyDZGI5FxQuSDouk58vylu9JTqIlvVwKSztmbkiwSK6wyLhzuforryvsSktpSMBLlrOsGzKcSt5X1gMJOGnTX0vasFwFGTdunOrMSfaJXAEpeOAsa9tkaKJ0iOQAJFMVS/BOOgxS6LFgNsvlkE64rGuyLKVAZUFyFUqCE9LJkYOnjKOXq0PSDlnfpBZHSWR702oSSWdAOmiynRV21VD+NnlOhizItONah1GGmDkWoJR1QLYx+S6k8ylj9zds2KACeVKUU7briiRBKFn+MhxPOkEFs4Kk4ybLQtotdTikIyntluVbXP0OWY9kOUonUbb9m2++WaWwS0dMthdZ1lrWVWlIJ04CZ/L9SLBL0sbluy04XbU8J1M6yzop66x0luVEQJukQTICJYiuBc1kHZF2SmdUhjM6kv2W1JGQDFIicm4SlJAAsvRzRGmCUtJ3kWOcZLped911apiJZDPISa0cy+6++264grIcN2TfKLV4pG8mAXvpY0lfS2gXGCu75tHl9n+lHyL7bAlIyX5aMmTkmFLUkHAhIwLkIpscc+T10n+W71n6X1q2jvSNpY6OZKfI8U2OV9LfLUvNTkdyDJL3lv6rtFeChrLMJYCgkdEJcoyS9U+OSxIUlP6kY+HxsrattOcAFUHqCckxtLD6i9KX1/rGckFILsRJmQP5G+V4K8HE0pDvTAKssoxkyJ58d1JzSZalY81IWbclACTLWfp10of9448/8kYCOLavtOcmFUG2Mfk+5G+QPn/Bi3+yHksfRepPSZ9c1m/p78g5kfw9xZHlIRd2pRyBbPOyXGWdkMwpeVwCYbK+lKYYvfT3tOwmOW+R8xkpTyHtl2CqRvYrsu3I6ABZl+RvktdJn1g7b5T1Ws4dZN8i25hk/klfU871HPuZ8l3Kd1iaC8FUDmWcrY+oVLRpSR1v3t7e9rZt29o//PDDfFOViuTkZDVlac2aNe0Wi8XeqFEjNQ269jqZOlWmo3/sscfy/Z5M39mpUyf1ezJ9uuM0xTJ1qkydLFPbV6tWTU2v6jiNsig47arYsmWLmlLU399f/e7VV19t//vvvy/5G2fNmqWmHZVpVuV9ZFpdx7+9sKlwyzINrUwDXXA6eG3q3rJOxVpwalW5yVS7MgWqfP7QoUPt69evL/R3HJdRZmamfezYsWoqe5mKVpaz/H/GjBn5ficlJcV+11132YODg9Xva9P/Ftd+7TltOToun02bNqmpamUdkveSKbALku9bpsH18vJS3/dzzz1nX758+SXvWVTbtOVd8HuTaYSvvPJKu4+Pj1peMh307t27871Gm8654PTNBaf0Ls6NN95o7927d6HLRLvJthEREaGmAH799dfVFOEFFdUWcfLkSTXdt/ztMiX1bbfdpqYVLmw7WLFihZqyWqYhbtiwoX327Nn2MWPGqO/AkUzd+/LLL9vr16+v2hcVFWUfN26cPSMjI9/rZDkPGDCgxKmkiyP7A3n/wqaPFh999JFaNjKduawH0m5ZXxMTE0v1/jJd8iuvvKL+btn+5W+Xz5OpqQtOI13ctN+yL3rggQfs4eHh6n1kf7J3795LptOWv6Fz587q+5D1S6Ygl+9Vpo8WMTEx9hEjRqjHZVuT76xLly5qqm5Hsl+TaeZfeOGFUv2dRKS/6dOnq32I7AMKU9TxQ45/sk+Q/a0c6x555JG8/o9G9jPasU1TcD9/ucctra9VEu1zynPcOHPmjDpuSH9D3sPxWCF/X8G/sTAl9ZsK+zuK65uVtf8rx5Hq1avbr7nmGvv7779vT0pKKnEZyfH3pptuUn1b+X25v/POO+379+/P93s//vijvXnz5qp/7Nh/Ka79BY+52vL56quv1HcQGRmpjkey3I8dO3bJ77/77rv2WrVqqWOs9I2kf1bYcbyothW2bpZ0DqCR95FjYkEFj61Fkf69vMeaNWsuWSaO35kct6UNd999t/3XX38t9L2KaouYM2eO+n1ZRrKtyt9e2HaQmpqq3iM0NFR95sCBA+379u1Tr3vzzTfLfG5SVL+ksP51ceT8Rl4v2116enq+5w4fPmx/8MEHVf9K+oPSdmmL9JVLa9WqVapfJf0W+b6lb92xY0e1jE6fPl2qvqNYtGiRvXXr1qod9erVs7/11lv2Tz75JN/+S5abbDt16tRR34es39dff71abzXff/+9Ol+U52R7k9c+/PDDl7RFzl9l2Re2DdPlM8g/5QlmETkryU6SKzmFzaZF5KwkG0qyZuTKasFZcJyFXMWWcfieOsGAs5JhAXIFUq7oFjWTFBERkaeTDDvJwNdqezkbyUaXjGfJ2tJr8gMqnHwv0k+XoaVU8VhTiojICchwMUlHruhU7PKSoYqOJBAls8jIAZmciwxvlXR6BqSIiIiK9sYbb6gSDoVNhKN3P0vIcD6paSTlFsh5SEkY6QfLEGqqHMyUIrfDTCmiyycBDtmWZHY+6bxJXQ+pX7B161anzeQiIiIicgVSH04KZ0sNJ6kTJXXI5Ca1RR0n+CHyBCx0TkREl5AinVJ0UgrgSyF/KfYuVxgZkCIiIiK6PDLZgRQwlyLqciFdJiuRiWak6DuRp2GmFBERERERERERVTnWlCIiIiIiIiIioirHoBQREREREREREVU51pQCYLPZcOrUKQQEBMBgMOjdHCIiInISUuUgOTlZTSMusyI5i9WrV+Odd95RhXJPnz6NBQsWYODAgXnPS42SZ599FgsXLkRsbCzq16+Pxx9/HMOHD897TUZGBsaMGYOvv/5aTWTQt29fzJgxA9WqVStVG9h/IiIiosvtQzEoBagOVVRUlN7NICIiIid14sQJ1K5dG84iNTUVbdq0wYMPPoibb775kudHjx6NlStX4osvvkC9evXw66+/4tFHH1UdwxtvvFG9ZtSoUfj555/x3XffISgoCCNHjlTv9ddff5WqDew/ERER0eX2oVjoHEBiYiKCg4PVwgoMDNS7OUREROQkkpKSVOAlISFBBW6ckWQpFcyUatmyJf773/9i/PjxeY916NAB/fr1w2uvvab6PhEREZg/fz5uvfVW9fzevXvRrFkzrF27Fl27di3xc9l/IiIiosvtQzFT6mJnTkiHip0qIiIiKsjVhqfJdOOLFi1SmVSSHbVq1Srs378fkydPVs/LsL/s7Gz06dMn73eaNm2qpiUvbVCK/SciIiK63D4Ug1JEREREbmbq1KkYNmyYSpc3m82qlsOsWbPQs2dP9fyZM2dgtVpVppMjqSclzxVG6k7JzfEKKBEREdHlcJ6KnURERERUYUGpdevWqWwpyYp69913MWLECPz222/lfs+JEyeq9HvtxnpSREREdLmYKUVERETkRtLT0/Hcc8+pOlMDBgxQj7Vu3Rrbtm3DpEmT1JC96tWrIysrS9V5cMyWOnv2rHquMOPGjVMF1AvWiiAiIiIqL2ZKEREREbkRqRUlt4LTL5tMJthstryi5xaLBStWrMh7ft++fTh+/Di6detW6Pt6eXnl1Y9iHSkiIiKqCMyUIiIiInIxKSkpOHjwYN7PR44cUZlQoaGhqlj5f/7zH4wdOxY+Pj6oW7cu/vjjD8ybNw/vvfeeer0MvxsyZIjKfJLfkQDTY489pgJSpSlyTkRERFQRGJQiIiIicjGbNm3C1VdfnfezNqzuvvvuw6effoqvv/5aDbcbPHgw4uLiVGDq9ddfx/Dhw/N+R2bik2yqW265RRUw79u3L2bMmKHL30NERESeyWC32+3wcFITQa4YJiYmMhWdiIiI8rCPUDQuGyIiIrrcfgJrShERERERERERUZVjUIqIiIiIiIiIiKocg1JERERERERERFTlGJQiIiIiIiIiIqIqx9n3iIiICsjJycGvv/6Kc+fOITIyEtdeey3MZh4yiYiIiIgqEnvYREREDubPn49ffvkFNpst32P9+/fHXXfdpWvbiIiIiIjcCYNSREREDsGnn376SU1fe9ttt6F9+/bYsmULvvvuO/W4YGCKiIiIiKhisKYUERHRxSF7kiElAampU6eiV69eCA4OVvfyszwuz8vriIiIiIjo8jFTioiICFA1pGTInmRIFawfJT/feuutmDNnjnqdDOUjIiIiovx++OEHREdHq/+HhISoDHOjkbkwVDQGpYiIiABV1FzIkL3CtGvXLt/riIiIiOhf8fHx+P777/M91r17dzRo0EC3NpHzY8iSiIgIULPsCakhVZitW7fmex0RERER5Q9KiR41/NC/bmC+x4iKwqAUERERgGuvvVall0tR84J1o+RnufInz8vriIiIiCg/LQAV7mNWN8fHiIrCoBQREdHFulFSKyoxMRGPPfYYVqxYgbi4OHUvP8vj8nzBelNERERE9G8AKtBiQpDVpP4vfSmi4rBnTUREdJEU4xQyy54UNddIhtT111+f9zwRERER5RcTE6Pug71MCPJiUIpKh0EpIiIiBxJ4uv3229Use1LUXGpIyZA9ZkgRERERlRyUCvU2IcBigsHhMaKisIdNRERUxFA+IiIiIiodCUCZDECg1QSjwaCypc6f56zFVDzWlCIiIiIiIiKiy3L+/Hk1dE8CUiLUy4S42Djk5ubq3TRyYgxKEREREREREVG5ZWZmqkLn4d7/DsaSGfhybTYO4aNiMShFREREREREROV29uxZdR/h829QKuJigOrMmTO6tYucH4NSRERERERERFRuWuBJsqM0WoCKQSkqDoNSRERERERERFRup0+fviRTSgtQac8RFYZBKSIiIiIiIiIqt+joaHVf3deS91ikjwUGh+eICsOgFBERERERERGVmwSeLEYDQrxMeY9ZTQaEeZsZlKJiMShFREREREREROVis9lwKjoa1XzNMBokN+pf8lhCQgJSUlJ0ax85NwaliIiIiIiIiKhczp07h8ysLNRwGLqn0YbznThxQoeWkStgUIqIiIiIiIiIyuX48ePqvqbfpUGpWhcf015DVBCDUkRERERERERULlrASQtAOarpf+GxY8eOVXm7yDUwKEVERERERERE5aIFnArLlIr0MasC6AxKUVEYlCIiIiIiIiKicjly5AiCvUzwt/47855GCp9LsEpqSuXk5OjSPnJuDEoRERERERERUZklJiYiLi4OtS8O0ytMlL9FBaROnjxZpW0j18CgFBERERERERGVK0tKRPlbi3xN7YvPaa8lcsSgFBERERERERGV2eHDh/OyoYqiPae9lsgRg1JEREREREREVGaHDh1S93WKyZSq7mdRxc611xI5YlCKiIiIiIiIiMrEbrerQFOYd+FFzjUmg0FlS504fhxZWVlV2kZyfgxKEREREREREVGZxMTEICkpqdgsKU2dACtybTYcPXq0StpGrkPXoNTq1atxww03oGbNmjAYDFi4cGGRrx0+fLh6zZQpU/I9LpX+Bw8ejMDAQAQHB2PIkCFISUmpgtYTEREROW8fas+ePbjxxhsRFBQEPz8/dOrUCcePH897PiMjAyNGjEBYWBj8/f1xyy234OzZs1X8lxARkas6ePBgXsCpJHUvvkb7HSKnCEqlpqaiTZs2mD59erGvW7BgAdatW6c6XgVJQGrXrl1Yvnw5fvrpJ9VJGzZsWCW2moiIiEhfJfWhZDhFjx490LRpU6xatQo7duzA+PHj4e3tnfeaUaNGYfHixfjuu+/wxx9/4NSpU7j55pur8K8gIiJXpgWY6gUyKEXlZ4aO+vXrp27FiY6OxmOPPYZly5ZhwIABl1wBXLp0KTZu3IiOHTuqx6ZOnYr+/ftj0qRJhQaxiIiIiFxdSX2o559/XvWH3n777bzHGjZsmPf/xMREzJkzB/Pnz0evXr3UY3PnzkWzZs3UhcCuXbtW8l9ARESuTgJMUi+qdimG74V4mRBoNeHAgQNV0jZyHU5dU8pms+Gee+7B2LFj0aJFi0ueX7t2rRqypwWkRJ8+fWA0GrF+/foi3zczM1ONfXW8EREREbkD6T/9/PPPaNy4Mfr27YvIyEh06dIl3xC/zZs3Izs7W/WbNJJVVadOHdW/IiIiKk5OTg6OHjmCWn5mNbNeSWSoed0AC2JjYxEfH18lbSTX4NRBqbfeegtmsxmPP/54oc+fOXNGdbQcyetDQ0PVc0WZOHGiqq+g3aKioiq87URERER6OHfunKqv+eabb+K6667Dr7/+ikGDBqmheTJMT0g/yWq1qot7jqpVq1ZkH4oX9YiISHPs2DFk5+SgbqBXqX+HQ/jIpYJScgXv/fffx6effqqiqhVp3LhxKm1du504caJC35+IiIhIz0wpcdNNN6m6UW3btsWzzz6L66+/HjNnziz3+/KiHhERabRhePVKUeRcUy/gQgCLQSlyiaDUmjVr1JU+SSOX7Ce5STR2zJgxqFevnnpN9erV1WsKphHKjHzyXFG8vLzUbH2ONyIiIiJ3EB4ervpNzZs3z/e41IvSZt+TflJWVhYSEhLyvUZm3yuqD8WLekREpNECS1r2U2lEBVgg6SYMSpFLBKWklpTMFLNt27a8mxQul/pSUvRcdOvWTXWmJKtKs3LlSnWFUGonEBEREXkaGZbXqVMn7Nu3L9/j+/fvR926ddX/O3ToAIvFghUrVuQ9L6+XoJX0rwrDi3pERKSRwJKfxYgwb1Opf8fLZERNPwsOHzqE3NzcSm0fuQ5dZ9+TegeOUdIjR46o4JPUhJIMqbCwsHyvl86TXL1r0qRJ3hU/qZUwdOhQlY4uBTtHjhyJO+64gzPvERERkdsqqQ8lF/H++9//omfPnrj66qvVbMWLFy/GqlWr1Otl+N2QIUMwevRo9TsSYJLZjiUgxZn3iIioOFJTUEYsNQ/1LnOpnToBVkSfSUV0dLQ6XhHpmim1adMmtGvXTt2EdIzk/xMmTCj1e3z55ZdqtpjevXurqY979OiBjz/+uBJbTURERKSvkvpQUthcLti9/fbbaNWqFWbPno3//e9/qp+kmTx5sqozdcstt6jglVz4++GHH3T7m4iIyDUcOnSozEP3NCx2TgUZ7Ha7HR5OIr1yxVDqIzAVnYiIiDTsIxSNy4aIyDN9//336iLG8JbhaBriXabfPZ2ajbe2nMVVV12FYcOGVVobyXX6CU5bU4qIiIiIiIiInMvhw4fVfZR/2TOlqvmaYTUZ8t6DiEEpIiIiIiIiIiqRDLQ6cuSwKnAuhc7LymgwoLafBdEnT6pZYIkYlCIiIiIiIiKiEsXFxSExMalcWVIa+d1cm03N+ErEoBQRERERERERlUhmexW1/S3lfo+oAEu+9yLPxqAUEREREREREZXo2LFj6r72ZWRK1fKz5nsv8mwMShERERERERFRibQhd7X8yp8pFeljhtloYFCKFAaliIiIiIiIiKhER48eRaDVhACrqdzvYTIaUN3XjBMnjsNms1Vo+8j1MChFRERERERERMXKyMjA+fPnUcPXfNnvVdPPgqysbJw9e7ZC2kaui0EpIiIiIiIiIipWdHS0uq9+GUP3NNV9LfnekzwXg1JEREREREREVLqg1MWA0uWQ4XuO70mei0EpIiIiIiIiIiplUOryh+9pga2TJ09e9nuRa2NQioiIiIiIiIiKdfr0aXUf6XP5mVLBXiZYjIa89yTPxaAUERERERERERXrzJkz8DUb4We5/DCC0WBAuLdJvafdbq+Q9pFrYlCKiIiIiIiIiIpks9nUTHnhPqYKe88IHzPS0tKQkpJSYe9JrodBKSIiIiIiIiIqUnx8PLKzsxHhffn1pDThPhfeS7KlyHMxKEVERERERERERTp//ry6D63AoFSo14X3iomJqbD3JNfDoBQRERERERERFUkLHFVoUMrblC/gRZ6JQSkiIiIiIiIiKjko5VVxNaVCmClFDEoRERERERERUXFiY2PVfXCFBqVM+d6bPBODUkRERERERERUpLi4uAoPSnmbjfAyGVURdfJcDEoRERERERERUZEkcORtuhBEqkhBVmNewIs8E4NSRERERERERFQkCRwFWis+fCCZV0lJScjJyanw9ybXwKAUERERERERERUqNzcXycnJCLJW3NA9TeDF90xMTKzw9ybXwKAUERERERERERVKMpnsdjsCKiFTKsBy4T0ZlPJcDEoRERERERERUaG0gFGApeIzpQKYKeXxGJQiIiIiIiIiouKDUpWQKaXVqUpISKjw9ybXwKAUEREREREREVV5ppT/xfeUIYLkmRiUIiIiIiIiIqJCaQGjysiU8r9YU4pBKc/FoBQRERERERERFZsppWU1VSQt+4o1pTwXg1JEREREREREVMLwvcrLlGJQynMxKEVERERERERExWdKXZwpryKZjAb4mo0cvufBGJQiIiIiIiIiokJJwMjHbITFaKiU95cMrETOvuexGJQiIiIiIiIioiIzpSpj6J4m0GpCckoKcnNzK+0zyHkxKEVEREREREREl7DZbJUelJJZ/ex2O4fweSgGpYiIiIiIiIjoEhKQkoBRkFfF15PSBF2sVRUfH19pn0HOi0EpIiIiIiIiIrqEFijSAkeVQQt4MSjlmRiUIiIiInIxq1evxg033ICaNWvCYDBg4cKFRb52+PDh6jVTpkzJ93hcXBwGDx6MwMBABAcHY8iQIUhJSamC1hMRkavQAkVS96myaO8txyXyPAxKEREREbmY1NRUtGnTBtOnTy/2dQsWLMC6detU8KogCUjt2rULy5cvx08//aQCXcOGDavEVhMRkauJiYlR9yGVOHxPe+/Y2NhK+wxyXma9G0BERO4hMzMTp06d0rsZVAIJTnh5eendDLpM/fr1U7fiREdH47HHHsOyZcswYMCAfM/t2bMHS5cuxcaNG9GxY0f12NSpU9G/f39MmjSp0CAWERF5blAq1LvyQgehXuZ8n0WehUEpIiKqEBKQev755/VuBpXg9ddfR/369fVuBlXBbEn33HMPxo4dixYtWlzy/Nq1a9WQPS0gJfr06QOj0Yj169dj0KBBhQae5abhLElERO7v/PnzlZ4pJbPvmQyGvM8iz8KgFNHFzvvevXuRkJCgOulNmzZVHXMiKj3JrJCAh7uQLJMZM2bg0UcfRa1ateAumAHjGd566y2YzWY8/vjjhT5/5swZREZG5ntMXh8aGqqeK8zEiRPx8ssvV0p7iYjIOZ07dw5WowH+lso7NzIaDAj1NqnPIs/DoBR5vA0bNuDLL7/MF5mPiIhQtTY6d+6sa9uIXIkMCXPHDBwJSLnj30Xua/PmzXj//fexZcsWVeC8oowbNw6jR4/OlykVFRVVYe9PRETOxW6348zp04jwMVfo8aQw8hm74xKRlpYGX1/fSv0sci5MBSF4ekBKOu7SqZarv5988om6l5/lcXmeiIjIlaxZs0Zdba5Tp47KfpLbsWPHMGbMGNSrV0+9pnr16pdckc7JyVEzH8lzRQWeZaY+xxsREbkvGUWSkZmpAkaVTfuMorJ1yX0xKEUePWRPMqTatWunrvw2atQI3t7e6l5+lsfleXkdERGRq5BaUjt27MC2bdvybjJsU+pLSdFz0a1bN3WyIVlVmpUrV6pjXpcuXXRsPbnqlPEy5FmyKojIfZw+fVrdR1ZBUEr7DE6a43k4fI88ltSQkiF7I0eOvKR+lPx844034qWXXlKva968uW7tJCIiKiglJQUHDx7M+/nIkSMq+CQ1oSRDKiwsLN/rLRaLyoBq0qSJ+rlZs2a47rrrMHToUMycORPZ2dnqeHjHHXew7hiVSXJyMkaMGKH+P3z4cPTs2VPvJhFRBTlx4oS6r+ZrqfTPqn7xM06ePFnpn0XOhZlS5LHkCrEoqh6G9rj2OiIiImexadMmldErN6Fl+E6YMKHU7yHZwDKxR+/evdG/f3/06NEDH3/8cSW2mtyRDPnUsEgxkXs5fvy4uq/lV/lBqRoXP0P7TPIczJQijyWz7GlXAGTIXlFXBrTXEREROYurrrqqTEOljh49esljklU1f/78Cm4ZeRopSqxJT0/XtS1EVLEkQGQ2GhDhW/lhA1+zESFeJlUDkTwLM6XIY8nVYZll78cff7ykbpT8vGjRIvW8vI6IiIiIig9KOf6fiFxbbm4uTpw4jhq+ZpgqeeY9jWRkSY06md2VPAeDUuSxpG7U4MGDsXXrVrz33nvYv3+/usIn9/KzPC7PF6w3RUREREQXpKamFvp/InJtMmokKysbdQKsVfaZ2mc51kwk98fhe+TROnfujCeeeELV1ZCi5hrJkJLH5XkiIiIiKpxjIEoK8BORezhw4IC6r1uFQal6DkGp9u3bV9nnkr50TQFZvXo1brjhBjXLi8FgwMKFC/Oek1lgnnnmGbRq1Qp+fn7qNffee+8lU0RKcUXJZgkMDFS1f4YMGcIDIpWJBJ4mT56MF154Qc08JPfyMwNSRERERCXPvlfY/4nItWnZSlqgqKoypQwOATHyDEa9r6y0adMG06dPv+Q5GZO+ZcsWjB8/Xt3/8MMP2LdvH2688cZ8r5OA1K5du7B8+XL89NNPKtA1bNiwKvwryB3IEL3mzZuje/fu6p5D9oiIiIhKllf7xWhgHRgiN7J37174WYyI8Km6wVXeZqOahU8CYjk5OVX2ueTBw/f69eunboUJCgpSgSZH06ZNU9krMgtAnTp1sGfPHixduhQbN25Ex44d1WumTp2qpjWeNGmSyq4iIiIiIqLKoQWizMEBSIlPVsWRTSaT3s0iostw7tw5nD9/Hm3CfdSIpqp0RZAXTp1KwaFDh9CkSZMq/WzSh0ulgyQmJqqNQobpibVr16r/awEp0adPH5Xlsn79eh1bSkRERETk/hISEmAwGmEO9IXdbme2FJEbkJFIolGQV5V/duNgr3xtIPfnMkGpjIwMVWPqzjvvVPWjxJkzZxAZGZnvdWazGaGhoeq5omRmZqoDpuONiIiIiIjKRqZvN/p6weTjnRekIiLXpgWEtABRVWoQ5KXqSjEo5TlcIiglRc9vv/12dfXlww8/vOz3mzhxohoeqN2ioqIqpJ1ERERERJ7CZrOpIJTRx0sFprRJiIjItbfrHdu3I8TLVKX1pDS+ZqMqeL5/3z5VZ5rcn9FVAlLHjh1TNaa0LClRvXp1Nd7VkRREk4OhPFeUcePGqaGA2u3EiROV+jcQEREREbkbmW1P1ZDy9VY3waAUkWuTme9SUlPRItS7yutJaeSzcyU4tmOHLp9PVcvoCgEp2TB+++03hIWF5Xu+W7du6urM5s2b8x5buXKliu526dKlyPf18vJSwS3HGxERERERlV5sbKy6N/l5q5vjY0TkmmTme9Ei1Ee3NrQM887XFnJvus6+l5KSoqZ71Bw5cgTbtm1TNaFq1KiBW2+9Va2IP/30k7oKo9WJkuetViuaNWuG6667DkOHDsXMmTNVEGvkyJG44447OPMeEREREVEliomJUfcmPx91c3yMiFyTJHxYjQZcoUM9KU0NX4saPrht61Y1EkrqRpP70jVTatOmTWjXrp26idGjR6v/T5gwAdHR0Vi0aBFOnjyJtm3bqiCVdvv777/z3uPLL79E06ZN0bt3b/Tv3x89evTAxx9/rONfRURERETk/mTKeCEBKaO3F2A0MChF5MKkrM2pU6fQPNQbFqM+Q/eEDBtsHeajhhGy4Ln70zXkeNVVV6ni5UUp7jmNZE3Nnz+/gltGRERERETF0Wq7mvx9YDAaVHCqYL1XInIda9euVfftIvQbuqdpG+GDP06lYP369WjTpo3ezSFPrSlFRERERETOH5QS5gBfVe81MzNT55YRUVlJQsj6detgNRnQLET/oFS9AKsawrdx4wY1hI/cF4NSRERERERUZmfPnlXD9owWc77glDasj4hcx+HDh3H6zBm0DPVWgSm9yRA+ydhKTU1TdafJfTEoRUREREREZSKTEJ07fw6mQN+8x0wBfur+9OnTOraMiMpj9erV6r5ztQvbsTPoFOmXr23knhiUIiIiIiKiMpGC5rZcmxqypzFfDFBpM2YTkWvIysrCX3/9hWAvExrrOOteQTX8LIjyt2Drli1ITEzUuzlUSRiUIiIiIiKiMtGyofIFpS7+n0EpIteyefNmpKWloVOkL4wG/YfuOZLMrVybDX/++afeTaFKwqAUERERERGVKyhlCvx3qI/J31cKwXD4HpGL+e2335xu6J6mQ4QvLEYDVq5YAZvNpndzqBIwKEVERERERGWiZUOZHYJSBpNRFTtnUIrIdZw4cQJ79uxB0xAvRPhcmLTAmfhajKrguRRh37Vrl97NoUrAoBQRoKLuu3fvxt9//63uGYUnIiIiKtqpU6cuGb6nfg70U7VfZCgQEblOllSPGv5wVv93sW3Lly/XuylUCZwvFEpUxTZs2IAvv/wy3/TFERERGDx4MDp37qxr24iIiIickWRDSVaUwWzK97gUO8+MvvB8w4YNdWsfEZUsNTUVa1avRoiXCc1DveGsogKsqBtgVbWv5JxNztXIfTBTiuDpAan3338fUVFRePnll/HJJ5+oe/lZHpfniYiIiOhfGRkZiIuLg6lAlpRjjSkO4SNyfitWrEBGZiZ61vR3ugLnBf2npj/sdjt++eUXvZtCFYxBKfJYMkRPMqTatWuH0aNHo1GjRvD29lb38rM8Ls9zKB8RERFR8fWkNNpjnIGPyLllZ2dj6dKl8DYZ0a268xU4L6hNhA9CvU34/fffkZKSondzqAIxKEUea+/evSr986abboLRmH9TkJ9vvPFG9by8joiIiIiQLwvKHFR0UEqrOUVEzumvv/5CQkICutfwhbfZ+cMCJoNBZUtlZWWxtpSbcf61j6iSyE5YyFC9wmiPa68jIiIiIoegVMClQSmjj5eqM8VMKSLnJSNBFi9apAI9PWsGwFV0re4HX7MRS5csUcOIyT0wKEUeKzg4OG8a1MJoj2uvIyIiIqJ/h+aZAi+tKWUwGFRdqdNnTqv6L0TkfGTG8dNnzqBrdV8Ee+WfrKAifL43FpO2nlX3FcnLZMRVtfyRnJLCbCk3wqAUeaymTZuqmRt+/PHHS+pGyc+LFi1Sz8vriIiIiOjfoJTBaITJ16fQ580BvsjMyERiYmKVt42IiifnOQt++EFlSfWJqpwsqbPpOTiZkq3uK5oUZZdsqZ8WL2a2lJtgUIo8ltSNGjx4MLZu3Yr33nsP+/fvR3p6urqXn+Vxeb5gvSkiIiIiTw9KGf19YDAaigxKaa8jIufNkgrxMsPVSP0rLVvq119/1bs5VAFcby0kqkCdO3fGE088oWbZe+mll/IelwwpeVyeJyIiIqIL0tLS1MxXXrUiinyN6WJQ6uzZs8w4J3KyGfe++/ZbmI2VlyVVFSRb6o9TKWpkS69eveDv7693k+gyMChFHk8CTx07dlSz7ElRc6khJR0oZkgRERER5Xfu3Ll8gaeSglJE5DykDtP5mBj0qu3vkllSjtlS10YFYMHhRFWKRUa3kOty3TWRqAJJAKp58+Z6N4OIiIjIJYJSZv/C60ldeM4332uJSH+S4bhgwQ+qHlOf2oFwdVfW8MfqU6lYtmwZrr32WjXShVwTU0GIiIiIiKhUzp8/r+5NFwNPhTH6eAFGA2JiYqqwZURUHMkoSk1NwzVRAfC1uH4YQIYgDqgbiJycHHz11Vd6N4cug+uvjUREREREVMVBqaIzpaQAusnPJ++1RKSvkydPYumSJQj3MeP/arpP/aW2ET6oF2DFunXrsHv3br2bQ+XEoBQRERGRi1m9ejVuuOEG1KxZEwaDAQsXLsxXyPaZZ55Bq1at4Ofnp15z77334tSpU/neIy4uTtXhCAwMVPUUhwwZooZ3EBVHy36SoFNx5Pn4+Hi1PhKRfux2Oz777DPk2my4uUGQyjByF0aDAbdcEQz5iz799FOVNUWuh0EpIiIiIheTmpqKNm3aYPr06YXOjrZlyxaMHz9e3f/www/Yt28fbrzxxnyvk4DUrl27VOHbn376SQW6hg0bVoV/Bbmi2NhYGK0WGK3Fl6Y1+XnnBT+JSD8bNmxQ+/qWod5oHlp8MNkVRflb0a26n8oGk+MZuR4WOiciIiJyMf369VO3wgQFBV3SMZ82bZqabfb48eOoU6cO9uzZg6VLl2Ljxo1qBloxdepU9O/fH5MmTVLZVUSFkZm7jBcDTqUJSkkQq1q1alXQMiIq7AKGZElJdtTABsFwVwPqBWJ7bDq++/ZbdOrUCeHh4Xo3icqAmVJEREREbi4xMVEN85NhemLt2rXq/1pASvTp00fNRrt+/XodW0rOLCMjA2mpqTD5liIo5XshI4PFzon0M3/+fCQkJOC6OgGqnpS78rOYMLB+EDIyMzFnzhw1ZJFcB4NSRERERG4eSJAaU3feeaeqHyXOnDmDyMjIfK8zm80IDQ1VzxUmMzMTSUlJ+W7kWSTryTELqjgcvkekLxmy9/vvv6O2vwVX1w6Au+sY6YumIV7Yvn07/vrrL72bQ2XAoBQRERGRm5Ii07fffru6avzhhx9e1ntNnDhRDQ3UblFRURXWTnINWoCpNJlS2hA/LZBFRFV7MWLWrI8hNc3vaBQCk8F9ipsXRbKBb78iBF4mI+bNm6cyhMk1MChFRERE5MYBqWPHjqkaU1qWlKhevTrOnTuX7/Uya5EEHeS5wowbN0518rXbiRMnKv1vIOeiBZhKVVPqYuCKmVJEVe/zzz/HuXPn0bt2AGr7W+EpQr3NuKFeoJpJ9uOPP+YwPhfBoBQRERGRmwakDhw4gN9++w1hYWH5nu/WrZuqM7J58+a8x1auXAmbzYYuXboU+p5eXl4qsOV4I8+i1YcqVaaUxaxm6WNNKaKqtWnTprxhe33reN5++soafmoY39atW7FixQq9m0OlwKAUERERkYuRq8Dbtm1TN3HkyBH1f5ldTwJSt956qzox+fLLL5Gbm6vqRMktKytLvb5Zs2a47rrrMHToUDVduNTfGDlyJO644w7OvEclD9/zK9208kZfL8TEMihFVFXkYsOsjz+GxWjAPU1C1ax7nkaG8d3VOBR+FiO++PxznDp1Su8mUQkYlCIiIiJyMRJwateunbqJ0aNHq/9PmDAB0dHRWLRoEU6ePIm2bduiRo0aebe///477z0kYNW0aVP07t0b/fv3R48ePdRwB6KinD9/vtSZUup1fj5IT0tHWlpaJbeMiCTTVWoHJqek4Kb6Qajma4GnCrSaVC2trOxsTJ06Ne+CDDkn950XkoiIiMhNXXXVVcXWyihNHQ2ZaU+mCycqS1BKAlIGU+mua5v8ffJ+r27dupXcOiLPtnDhQuzcuRMtw7zVEDZP1yrMBz1q+OHPY8dU4fOHHnpI7yZREZgpRUREREREJWZhxMTGlqrIucbk76vuCxbVJ6KKJcGo//3vfwjzNuOuRqFqCBsBAxsEI8rfomom/vnnn3o3h4rAoBQRERERERVLsp1subkwB5Y+A8MccCEoJfXMiKjyar1NnzYNUj7q/qah8LXwFF8jNbXuaxoGH7MRs2fPVsPayflwjSUiIiIiomJpxYLLFJQKuvDa06dPV1q7iDyZ1EqaPHkykpKTMah+EKICrHo3yemE+5hxV+MQtazeffddNVEIORcGpYiIiIiIqFhaYMkUeCH7qdQ1pQwGzn5FVAmkdqBk/xw6dAidq/myjlQJ9aWujQrA2bNn8cEHH6hZacl5MChFRERERETFOn78uLq3BAeU+ncMRqPKlpLflZpURFRxfvnlF1UnqV6AFbdfEcI6UiW4rm6gKgL/zz//cJIPJ8OgFBERERERFevo0aMwWMwwXawTVVqW0EBkZGSomlREVDG2bdumAivBXiY82DxM1U6i4hkNBtzdOBTVfS1YsmQJVq1apXeT6CIGpYiIiIiIqEiZmZk4GX0S5pCAMmdjmEMC1f3hw4crqXVEnhcg/uD992E2AEOahSHQatK7SS7D22zEQ83D4GcxYs7s2WrWQtIfg1JERERERFSkgwcPwpZrgzUiuMy/q/3O3r17K6FlRJ4lNjYW77z9tgoU39s0lIXNy1n4XAJTBtgxZcrkvKHJpB8GpYiIiIiIqEh79uxR99ZqoWX+XUtYIAxmE4NSRJcpLS0Nb731FuITEjCoQbAq3k3lUz/QC3c3DkF6eoYK8sXHx+vdJI/GoBQRERERERVJCgPD8G/WU1lIsXNLRDBOnDiBhISESmkfkbvLzs7G5MmTcfLkSfynlj961vLXu0kur22EL26sH4TYuDgV7EtNTdW7SR6LQSkiIiIiIipUYmIiDhw4AGtkKIxWS7new7t2pLrfsmVLBbeOyP3JzJXTpk3Drl270DbcBzfVD9K7SW7j6lr++L+a/moI36RJk5CVlaV3kzwSg1JERERERFQoCSTZ7XZ4R10ILJWH9rubNm2qwJYRuT/Z9ubMmYONGzeicbAX7m4SqmaRo4ohEzcMahCE9hE+2LdvH6ZOnYrc3Fy9m+VxGJQiIiIiIqJC/f333+reK6paud/D5OcDS2igmukqKSmpAltH5N6+/fZb/P7774jyt+DBZmEwGxmQqmgS5LurcSiahnhh8+bNmD17tspOo6rDoBQREREREV3i7NmzasiQtXoozP6XV1TZp2EtlYGwZs2aCmsfkTtbtGgRfvzxR0T4mPFwi3B4m3nqXlkk2PdAszDUDbDijz/+wOeff66y1KhqcM0mIiIiIqJLSIaG8L2i9mW/l0/9mjCYjFi5ciVP9ohKsGzZMnz99dcI8TLh0Vbh8Lea9G6S2/MyGTGsRThq+lnU8v/mm2/0bpLHYFCKiIiIiIjyycjIwIoVK2D0ssK7TvmH7mmMXhZ4162O06dPY/v27RXSRiJ3tGrVKnz22WcItJowolUEQrzMejfJY/hZjHikZTgifcwqU23hwoV6N8kjMChFRERERET5/Pbbb2qKdL9mdWEwVUyWhl+L+up+wYIFzJYiKqKG26xZs1Rw5NGW4Qj3YUCqqgVYJTstAmHeZlXT65dfftG7SW6PQSkiIiIiIsoj06L/9PPPMFrN8G1Sp8Le1xIcoLKuDhw4gN27d1fY+xK5g3Xr1mHGjBnwNhlUtk51P4veTfJYwReHTcr9F198oYbzkZsGpVavXo0bbrgBNWvWVNMxFkyPkysoEyZMQI0aNeDj44M+ffqog5ijuLg4DB48GIGBgQgODsaQIUOQkpJSxX8JEREREZF7+Omnn5CUmAjfpnVhtFbsibF/q4bq/ssvv+QMV0QXbdiwAdOmTYOXESogVdvfqneTPJ5kSo1oFY4gq0kNp1y+fLneTXJbugalJCW4TZs2mD59eqHPv/322/jggw8wc+ZMrF+/Hn5+fujbt68a466RgJTMCiIriRxAJdA1bNiwKvwriIiIiIjcQ2xsLH5ctAgmXy/4Nb8w3K4iWUID1Ux8R48eVbNcEXm6TZs2YeoHH8BqBIa3DEedAAaknEWEj0VlTEl9r7lz56o6e+RmQal+/frhtddew6BBgy55TrKkpkyZghdeeAE33XQTWrdujXnz5uHUqVN5GVV79uzB0qVLMXv2bHTp0gU9evTA1KlT1UwF8joiIiIiIiq9+fPnIzsrC/7tmsBoqZx6NgFtG8FgMavZrTjCgTzZ5s2b8f7778NsAB5uEYa6DEg5nWq+FpUxJbWm5syZkzcrKXlATakjR47gzJkzasieJigoSAWf1q5dq36Wexmy17Fjx7zXyOuNRqPKrCIiIiIiotKfIEv/2hIRDJ/6NSrtc0y+3moYX1JSkqrXQuSpGVKShGGGXQWk6gd66d0kKiYwJYXn/S0mVYh+5cqVejfJrThtOX8JSIlq1fJPQSs/a8/JfWRkZL7nzWYzQkND815TmMzMTHXTyAGRyk6WITPSnJ/UbPPy4kGOiIiIipacnKxGHxhMRgR3a6nqvVYmmdUv49gZVXpDLjq3a9euUj+PyNkCUipD6mJAqkEQ++rOroafBSNbhWPazhi1rxS9evXSu1luwWmDUpVp4sSJePnll/VuhsuTgNTzzz+vdzOoBK+//jrq16/4mhBERETkHqRshtRLSUxMRECHJjAH+Vf6ZxqMRgR1b4XYn//Gx7Nm4a0331QTFxG5u40bN+IDhyF7DEi5juoXA1PT/7kQmJJ9Z+/evfVulstz2qBU9erV1f3Zs2fV7Hsa+blt27Z5rzl37ly+38vJyVEz8mm/X5hx48Zh9OjR+TKloqKiKuGvcP8MHAl4uIvo6Gg1Deujjz6KWrVqwZ2+JyIiIqKiyFAUmY7eGhkCv6b1quxzLcH+8G/XCImb9+HDDz/E2LFjVRkOIneeZU+KmktAanhLDtlz1cDUiJYXAlNSY0oCU44lh8iNglKS2SGBJalwrwWhJHgktaIeeeQR9XO3bt2QkJCgxr936NAh76Aq08tKGnBRZCgThzNdPlmG7piBIwEpd/y7iIiIiAo6duyYmu7c6GVFcI82MBgrd9heQX7N6iHrTBy2b9+uZtK+8cYbq/TziaqKnMfKpFwWBqTcJ2NqZww++eQTFX+49tpr9W6Wy9I1KCWzbRw8eDBfcfNt27apmlB16tTBk08+qWbna9SokQoSjB8/XmV9DBw4UL2+WbNmuO666zB06FDMnDkT2dnZGDlyJO644w5mhxCRS4iJiVF1PMg5sycd78n5BAQEIDw8XO9mELms1NRUTHn/fTXSIKRnB5j8vKu8DVK7KvjKVoj56W98++23uOKKK9C8efMqbwdRZZJMxGnTpsFqBIa3CEM9BqTcZlY+CUx9+umnKmOqb9++ejfLJZn1LvB29dVX5/2sDam777771Bf79NNPq4PlsGHDVEZUjx49sHTpUnh7/3vA/PLLL1UgSsZySrrvLbfcgg8++ECXv4eIqKwBqafGjEFWdrbeTaFiyLBeck5WiwWT3n2XgSmicpAr+5K1cfbMGfi3bADvWhG6tUVlafVsg7hfN6jZyKQ8RESEfu0hqkh///236kt4SUCqZTjqBlj1bhJVYGBKy5iSjFPZr/br10/vZrkcXYNSV111lYooFnfl5JVXXlG3okhW1fz58yuphURElUcypCQgdXeTUFTzddrR1ERO6WxaDr7YF6e2IwaliMru66+/xo4dO+BVOwL+bRvp3RxYI0IQ2LkFEtf9o4LNL7/0Ur4L0USu6K+//lIBKW+TAY+0DEcdBqTcTqQEplpfmJXv888/V/GN/v37690sl8KzICIinUlAKsqfnRQiIqoaf/zxh6rfJLPsBV/ZRl0Idga+jWojOyEZJ/Yew/Tp0zFq1CgWPieX9eeff6oC/j4SkGoVzr6eG4vwkYypCEzfeR5ffPGFCkwNGDBA72a5DO7liYiIiIg8xD///INZs2apIXMhV7WD0epc16gDOzSBtUaYmshIynQQuaI1a9ZcCEiZDXiUASmPEOFjVoGpYC+T2ndJ4J9Kh0EpIiIiIiIPcPLkSUyePBl2A1RAyhzoB2djMBoR0rMtzMH+WLJkCX799Ve9m0RU5oCUTMKlAlItw1GbASmPEX4xMBXiZVIlhhYvXqx3k1wCg1JERERERG4uPj4eb731FtLT0xHUvRWskSFwVkarBaG9OsDk46WKB8vkSESuMmRPC0iNYEDKcwNTrS8Epr766itmTJUCg1JERERERG4sLS1NBaRiY2MR0K4xfOrVgLMz+fkg5Or2gNmEqdOm4sCBA3o3iah0NaQuBqRqMSDlscK8/w1MScbUzz//rHeTnBqDUkREREREbionJwdTpkzB8ePH4dukDvxa1IersIQFqaF82Tk5eOedd3D69Gm9m0RU5Cx7eTWkGJCii4GpEQ41phiYKhqDUkREREQuZvXq1bjhhhtQs2ZNNXPawoUL8z0vM/9MmDABNWrUgI+PD/r06XNJpklcXBwGDx6MwMBABAcHY8iQIUhJSaniv4Qqk81mw0cffaSKm3vXqYbAjs2cZqa90vKqGY6gri3Vuvnmm28iISFB7yYR5bN27VrMmDFDzbLHGlJUWI0pLTAldfLoUgxKEREREbmY1NRUtGnTBtOnTy/0+bfffhsffPCBqm2yfv16+Pn5oW/fvsjIyMh7jQSkdu3aheXLl6uaFxLoGjZsWBX+FVTZvvnmG5XBYYkIRvCVrWEwulZASuPbsBYC2jbC+fPn1botdbGInMGGDRvUftjbZMAjDEhRMYGpIC8TPv/8c3XMpfwYlCIiIiJyMf369cNrr72GQYMGXfKcZEnJcK0XXngBN910E1q3bo158+bh1KlTeRlVe/bswdKlSzF79mx06dIFPXr0wNSpU/H111+r15HrW7ZsmZr5yRzkh9Cr28NgNsGV+bVsAN/GUTh69KiaQVCGJRLpSQrwT/3gA1iNwPCW4YgKYECKig5MSZ2xQKsJc+fOxYoVK/RuklNhUIqIiIjIjRw5cgRnzpxRQ/Y0QUFBKvgkw0yE3MuQvY4dO+a9Rl5vNBpVZhW5fvaGBCJNvl4I7d0RRi/XP1mWYYeBnZrDKypSDUecNWuWCsAS6WHr1q14//33YTYAD7cIQ10GpKgEkb4WPNoqHP4WE+bMmYM//vhD7yY5DbPeDSAiIiKiiiMBKVGtWrV8j8vP2nNyHxkZme95s9mM0NDQvNcUlJmZqW6apKSkSmg9Xa79+/er4USSGRVydQc1i11li1+zHTlJqTAH+iHk/9pU2ufI8MOQHm0Q+9tGrFmzBuHh4bjtttsq7fOICiPDnqdMngwTbBjWIhz1A730bpJTeGvzGaRk29T/U3Mu3J9Kzcb4dafgbzHimQ7V4emq+1owolU4pu08j48//hhWqxXdunWDp2OmFBERERGVaOLEiSrjSrtFRUXp3SQqQGanmzRpErJzcxDcsy0soYFV8rkSkMqJS1L3lU2CbaFXtYc5wBcLFizA77//XumfSeQY9JVtzG7LxUPNwtAwiAEpjQSkki/ebBeTGOVeftaCVQTU8LOo+mNSh2zG9OnYvHkzPB2DUkRERERupHr1C1ejz549m+9x+Vl7Tu7PnTuX73mp0SMz8mmvKWjcuHFITEzMu504caLS/gYqu+TkZLz9zjtqlrqgri3UrHXuyuhtRYgMS/S2qmEwMpyPqCqGRr/11pvIzsrEA81C0TjEW+8mkYuSgvgy7FOGf77//hTs3LkTnoxBKSIiIiI3Ur9+fRVYciykKkPtpFaUNkxA7hMSEvJdoV25ciVsNpuqPVUYLy8vBAYG5ruRc5CAohT/PnvmDPxbNYRvw9pwd5IpFXJVO9gNUH97dHS03k0iN3by5EmVLZqRnoF7m4SiRWjlD4sl91Yv0AtDm4fBYLPh3UmTsG/fPngqBqWIiIiIXIxkw2zbtk3dtCv48v/jx4+rgtBPPvmkmp1v0aJF6grsvffei5o1a2LgwIHq9c2aNcN1112HoUOHqqLYf/31F0aOHIk77rhDvY5chxT7llkU9+7dC+961eHf5gp4CmtECIK6t0J6ejrefvtt1jmjSnH+/HkVkJL97p2NQ9A2wlfvJpGbuCLYCw82C0NuTg7eefttHDt2DJ6IQSkiIiIiF5yKvF27duomRo8erf4/YcIE9fPTTz+Nxx57DMOGDUOnTp3UydTSpUvh7f3vcJMvv/wSTZs2Re/evdG/f3/06NFDFV4l1/LLL79g9erVsIQHIbh7KxWU9CQ+9WqoQJwEDmQ2NMkaI6ooklH6xhtvID4+HoMaBKFzNT+9m0RuplmoN+5uEqKC6xL8lNqAnoaz7xERERG5mKuuukplyBRFAhOvvPKKuhVFZtqbP39+JbWQqsL27dvVd2jy9UbIVe1hMJngiWTIYk5CCvbs2YN58+bhwQcf1LtJ5AZSU1Px5psTVT2+vnUC8J9aAXo3idxUuwhfZOTY8M3BC0HQl156CWFhYfAUzJQiIiIiInIxcjX9gw8+AIwGVVvJ5OO5s4BJEFayxGS2wd9++y1fPTWi8sjKylKz7B0/fgI9a/rjujqsoUeVq1sNf9xQLwixsbF46803VYazpyhzUEquyh04cAC7du1ieiwRERFRKbEPRRUlIyMD702erIZ7BHVrCUtYEDydwWxSwTmZke/TTz9V2xpReeTm5mLatGmq8HSHCB8MbBDkccNiSR+9owJwdS1/nIyOVkFRCY56gjIFpaSIZuvWrVX9Ablv2LChqmlAREREREVjH4oqMrgptb+iT56EX7O68KnPwvQak58Pgv+vDXJtNkyZMgWJiYl6N4lccPuaO3eu2j83CfbCnY1DYWRAiqrQDfWD0DHSF/v378fUqVNVkNTdlSkoNXbsWHVl74svvsD333+P2rVr4+GHH6681hERERG5AfahqKIsWbIE69atgzUyBAHtm+jdHKfjVT0MAe0bq8LUMrzRE07oqOL873//w8qVK1HH34IHm4fBbGRAiqqW0WDAnY1C0DTEC5s3b8Ynn3xSbA1Jjyt0/ueff6qOlMzOIrp27ao6VVIEzs+PMxEQERERFYZ9KKoIcuVcFTb38UJwz7YwGFketjB+zeohOyZRFT7/7rvvcMcdd+jdJHIBq1atwg8//IBwHzOGtQiHl4nbF+nDZDTggWZhmL7jPH7//XdERERg4MCBcFdl2tLOnTuHRo0a5f1co0YN+Pj4qMeJiIiIqHDsQ9HlSkpKwvsffACb3a6GqHlyYfOSSP2foK4tYQrwxaJFi7B161a9m0QuMJPl7Nmz4WcxYniLcPhbPXMmS3IeXiYjhrYIR5i3Gd9++626uOWujGXdwUsVeDkoajej0Yjk5OR8jxERERHRv9iHosths9kwY8YMxMfFIaBdI1irherdJKdntJoR8p92MJiMmD5jBs6fP693k8hJHT16FO9PmQIT7BjaPExlShE5gwCrCcNahMHXYsRHH32kJkqBpwelZCxj48aNERISkneTDla7du3U/4ODg9U9EREREf2LfSi6HJLts2PHDnjVjoBf8/p6N8dlWEICENi5OdJSU1V9Kc56SQVJ7bF33nkHmZmZuKdJKOoFMgORnEs1XwseahYGg92Gye+9h+joaLibMoWBZTwjEREREZUN+1BUXlpdJDWzXPdWnJq+jHwa1kLW2TgcOnQIX3/9Ne6++269m0ROQgJRkyZNUoGpgQ2C0DrcR+8mERWqQZAX7mocgnl74/DOO2/jlVdeRWBgIDwyKCVX84iIiIiobNiHovJITEzEB1Onwm6AqiNl9LLq3SSXI0E8yZbKjk3EL7/8giZNmqBTp056N4ucYEjshx9+iCNHjqBbdT/8p6a/3k0iKlb7CF+cS8vB0uPnMWXKFDz33HMwm91jqGmZ/gpJLS/N1RlOvUpERET0L/ahqDwnzdOmTUNiQgICOzaFNSJY7ya5LKPFjOCe7RC7ZC1mzpyJOnXqoFq1ano3i3Qks6Fu2LABjYK8cGvD0u2fifTWt04AzqVnY8vevZgzZw6GDRvmFutuuYfvSW2E/v37q1kKatWqVRltIyIiInIL7ENReU6apaitd51q8G1aV+/muDxLsD8Cu7RA4l878P777+Oll16C1crMM0+0bt06LFy4EBE+ZtzfLAwmo+uf1JNnMBgMuLNxKGIzzuOPP/5QAfZ+/frBo4JS//nPf/L9bDKZ0LVrVzRo0KCi20VERETkNtiHorLYsmULfvzxR5gDfBHUraVbXAl3Br4NaiL7XByOHjiKzz77DEOHDtW7SVTFjh8/rrLlvM1GPNQ8DH6WMs37RaQ7i9GAB5uH4b2t5/Dll1+qwFSLFi3gyrgVEhERERE5idOnT2P69OmA0Yjgnm1htFr0bpJbCezUDJawQJW9uGLFCr2bQ1UoOTkZ7777LrKysnB34xA1qxmRKwqymvBAs1AYYFeZn+fPn4crY1CKiIiIiMgJZGRk4L3Jk5Geno6gbi1gCXWf2ZWchcFkQsh/2sHobcWnn36K/fv3690kqsIabXLyfl2dQLQM40x75NrqBXrhtobBSElJwXvvvaeCrR4blGI6MREREVHZsQ9FBU+aJUMq+uRJ+DWrB5/6NfVuktsy+fmo2QxzbTZMnjwZMTExejeJKtmCBQuwc+dOtAzzxrV1AvRuDlGF6FrdD1fW8MOxY8fUkGSPqCl18803X3I1Z/jw4fDz88v3+A8//FAxrSMiIiJyA+xDUUm+/vprbN68GV41wxHQvrHezXF7XtXD1FC+xA278c6kSXjpxRfh48PsGXckwSjZt4Z5m3FX41AYeUGA3MigBsE4npylhiQ3a9YMPXr0gFsHpYKCgvL9fPfdd1d0e4iIiIjcDvtQVBw5mfjpp59gDvJXGTwGIytsVAW/JnWQk5iCE/uOq6FdY8aMgZHL3q3Ex8dj+rRpkAn27m8aCl8zv19yL2ajQc0iOWnrOTWrb/369V1uZt8yBaXmzp1beS0hIiIiclPsQ1FRJDtqzpw5qsZRyNXtWdi8igV2bIrc5DRs3bpVfQ8PPfQQh9a6WR2ppORk3NowGFEBVr2bRFQpwrzNuLNRCD7ZE6sKn7/22muwWl1nfWeomIiIiIhIB3v37sUHH3wAu9GgAlLmAF+9m+RxDBdnOdRm5Pvuu+/0bhJVkCVLlmDPnj1oHeaj6u4QubPW4T7oWdMfJ0+exDfffANXwqAUEREREVEVk8K077zzDnJyc9RscNbwYL2b5LGMFjNCenWEOdAPCxcuxC+//KJ3k+gyHT9+HN988zUCrSb8t1Ews9/II1xfLwjVfM0qIPvPP//AVTAoRURERERUxSfMr7/+OtLT0xF0ZWtV3Jz0ZfK2IrRPR5h8vfHFF19g2bJlejeJyik7O1vNZJmTk4s7G4fAz2LSu0lEVcJqMuDuJlLMH5g5cyZSUlLgChiUIiIiIiKqwoCU1PuQk4Wg7q3gU6+G3k2ii0x+Pgi9phNMvl5qenUGplzTjz/+iBMnTqBHDT80C/HWuzlEVSrK34p+dQMRFxeH+fPnwxUwKEVEREREVAWOHj2aLyDl29C1ZkjyBDKEL/SaznmBqaVLl+rdJCqD6OhoFZQK8TLhhvr5Zz0l8hS9agegtp8Fq1atUnXVnB2DUkRERERElUxODF559VUGpFwsMDVv3jx8//33sNvtejeLSjHb3uzZs5Gbm6tm2/My8VSXPJPJYMDtjUIgldRkVlEZ0urMuKUSEREREVWiTZs2YeLEicjMzFQzvTEg5SKBqb5d1IyIP/zwAz799FMV9CDntXr1auzbtw9twn3QIsxH7+YQ6apOgBX/V9Mfp06dwuLFi+HMGJQiIiIiIqokv//+OyZPnoxcAxDSqwN86lbXu0lUSmZ/XxWYsoQGYvny5Zg6dSqysrL0bhYVIiMjA998843Kjrq5AWeyJBL96wYiyGpSQamEhAQ4KwaliIiIiIgqmGTVfPXVV5g1axYMVosqoO1VI0zvZlEZmXy81FA+a7VQrF+/Xs2amJiYqHezqIAlS5ao7+XqWv4I8uJse0TC22zEdXUDVZauZHw6KwaliIiIiIgqkJwAvP/+++rqtDnID2H9usIaxqLLrspoNSO0d0f4NKyFAwcOYPz48Th58qTezaKLkpKS1LYWYDXh6tr+ejeHyKl0ruaLar5mrFy5EqdPn4YzYlCKiIiIiKiCxMbG4pVXXsHGjRthrRGGsOu6qrpE5NoMJiOCurVEQLvGiImJwYQXX8TWrVv1bhYBKiAlw/f6RgWwuDlRIUXPb6gXpLJ3//e//8EZcaslIiIiIqqgGfaee+45HDlyBL6NohDaqwOMVovezaIKYjAY4N+ygSpWn5mVhUmTJmHhwoWcmU9HaWlpWLFiBYK9TOhW3U/v5hA5pRah3qjtb8G6devUhRNnw6AUEREREdFlkKDEsmXLVL2hlNQUBHZpjqCuLWAwsqvtjqRYfdh1XWD088a3336rhmpKpg7pM+OeLPseNfxgMhr0bg6R0wbUr6rlr7Klfv31VzgbHimJiIiIiC6jftRHH32Ezz77DFAFzTvDr3EdvZtFlUxm5Avv1w3W6qHYsGGDqjMlU69T1ZET7GVLl8JiNKBbddaSIipO23BfBFpNWLlyhdMF0RmUIiIiIiIqhzNnzqjaQpKtYQkPQviAbrBGhujdLKoiRm+rKoDu16weoqOj8fzzz6sZ+qhq7N69G2fPnUPHSF/4WXhaS1Qcs9GAK2v4ITU1TQXSnYlTb725ubnqqkP9+vXh4+ODhg0b4tVXX803blv+P2HCBNSoUUO9pk+fPmpWDCIiIiKiyiKFzJ97/nmcOH4cvk3qIOzaLjD5euvdLKpiMkQzsGNTVWcqy5arhvJ98cUXyMnJ0btpHrENCglKEVHJOkT45tt2nIVTB6XeeustfPjhh5g2bZoqHCk/v/3225g6dWrea+TnDz74ADNnzlRXJvz8/NC3b1+nS0kjIiIiItcnwYb58+dj8uTJyMzOQnCP1gjq3FzNzkaeXWcqvF9XmIP98csvv+C1115DXFyc3s1y66F7mzZthL/FiPqBVr2bQ+QSwn3MqOFrwY4dO5wqXuLUR8+///4bN910EwYMGIB69erh1ltvxbXXXpuXbiZZUlOmTMELL7ygXte6dWvMmzdPjeeWmTCIiIiIPBGzzSuHBBmkmPlPP/0Ec5Afwvp1hU/9mno3i5yEOcgfYdd1hXf9Gti/fz/GjRuHnTt36t0st3To0CHExyegZag3jAYWOCcqrdbh3sjOzlaBKWdhhhPr3r07Pv74Y7VTb9y4MbZv344///wT7733nnpeptuVsfzSidIEBQWhS5cuWLt2Le64444iC1LKTZOUlFQFfw0RUeHOpmXr3QQil8PtpnTZ5lJ8u0WLFti0aRMeeOAB1U96/PHH82Wby2skeCVBLMk2lzot3t4chlbQP//8o7L1k5OT4V2vhppdz2hx6q406UDWieArWyM9MhRJm/bgzTffxKBBg3DzzTfDyNkYK8zevXvVfbNQH72bQuRSmoV4Y9nxZOzbtw+dO3eGM3DqI+mzzz6rAkZNmzaFyWRSV/3k6tTgwYPV8xKQEtWqVcv3e/Kz9lxhJk6ciJdffrmSW09EVDpf7IvXuwlE5GYcs82FZJx/9dVXRWabC8k2lz6UZJsXdWHPU4cJ/fjjj/j+++9lXm0EdmkO30ZRaoptosLIuuHbOAqWsEDEr96OH374QWUhjhgxAoGBgXo3zy0cP35c3df2t+jdFCKXUtPPAjl6HTt2DM7CqYNS3377Lb788ks1bl+u8m3btg1PPvkkatasifvuu6/c7yuptKNHj877WQJfUVFRFdRqIqKyubtJCKr5slNFVNZMKQZ0qzbb3BMzzVNSUjBjxgzVBzX5+yCkZ1tYwoL0bha5CFlXZEbGhL92qmF8zz33nDqXueKKK/RumsuTE2pvkxGhXia9m0IOikr8mPQaE0KchdVkRISPGcePHVMXqJzhAotTB6XGjh2rsqW0jlGrVq3UDkgynSQoVb16dfX42bNnVT0Ejfzctm3bIt/Xy8tL3YiInIEEpKL8WaSTiJw729zTMs0PHz6MyVOmIDYmBl61IhB8ZSsYvbivprIxWi0IuaodUncfQdzWA2obuvvuu1WdXGc4GXTVyQZOnzqFOv5mLkOicmZLbYtJRUJCAkJCQqA3pw5KpaWlXTL2WjpWkkYtpP6BBKZWrFiRF4SSDpjMwvfII4/o0mYiIiIid8w296RM81WrVuGTTz5BTm4uAto2gl/LBjz5pXKTdce/RQOVOZWwZoeq4yZBzyFDhsBqZaCzrGTWsFybDYFWZkk5mxdffLHQxwMsrKfmTAIvbjtSI5FBqRLccMMN6qpenTp1VIdq69atKu38wQcfzNvBSwdLplxt1KhRXpFO6XANHDhQ7+YTERERuU22uSdkmksGxueff47ly5fD6GVB6H86wKtmuN7NIjfhVT1MDeeLX70Na9aswcmTJzFq1CiEh3MdK4usrCx1bzEyUExUHtq2o21LenPqkKXMcHLrrbfi0UcfRbNmzfDUU0/h4YcfVlMaa55++mk89thjGDZsGDp16qTG/i9dupSzxhAREZHHKku2uUbLNu/WrRs8UWJiIt544w0VkDKHBCCsXzcGpKjCmXy9EXZNZ1UsX2q7Pf/889izZ4/ezXIpDEoRXR5t28nOdo6ZjJ06UyogIEDNDCO3oki21CuvvKJuRERERMRs87KSjJW3334bMTEx8K5XA8HdWsJg5tAgqhwGkxFBXVuo2fmSNuxWwVAZynfVVVfp3TSXIDXyBENSROWjjUbXtiW9OXVQioiIiIjKl20uQSbJNj937pwKNkm2+YQJE/Jlm6empqpscyl22qNHD4/MNpdZ0aSgeUZ6OutHUZWSbClzkD/iV21Vs2XKJAO33377JVmOlJ9WAycxyzlOqIlcTWLmhW3HGepJCQaliIiIiNwMs81LZ+XKlaqgud0ABPdsC5+6F2ptEVUVa2QIwvp1RfzKzVi0aJGq6yYTNrEAetF8fX3h6+OD+MwcvZtC5JLiLwalwsLC4AwYhiciIiIij2K327Fw4ULMnj0bsJoRek1nBqRIN+YAX4Rd1xXWaqGqrpsMJU1PT9e7WU4tPCIi78SaiMpGArr+/v5OkxnNoBQREREReVRAav78+fj2229h8vdBWN8usEYE690sl3N+8Z84+91KdcuJT1aPyb38LM9R2ajZHnt3hHfd6ti9e7eqCSfTtVPhZEhyeo4NMenMliIqi4wcG86l56htyFkwKEVEREREHkFmH5TsqJ9//hnmYH8VkDIH+undLJdky8jKu8Fuv/Cg3f7vY1SuAujBPdrA54raOHz4MF5++WXEx8fr3Syn1KpVK3W/Nz5D76YQuZT9CZmw2YHWrVvDWTAoRUREREQekSEl9aN+//13WMKCEHZtZ5h8nWPoApHGYDSomfn8WtTHqVOnVMZUYmKi3s1yOtoJNYNSRGWjbTNaYNcZMChFRERERG4fkJo3b54qbG4JC0Ron44werGQNDknmYQgoF3jvMDUGxMnIiUlRe9mORUp0CzDjw4kZiIr16Z3c4hcgs1ux574DPj5+aJhw4ZwFpx9r4rFxMRwfLiTio6OzndPzjmbVHh4uN7NICIiF/PNN99g2bJlMIcEqLo9RqtF7yYRlSowZc+14cTeY5j45kSMf2G80xQmdgbdu3fH999/j03n0tC9hr/ezSFyenviM9QEAVd37wKj0XnykxiUquKA1JinnkJ2FsfZO7MZM2bo3QQqgsVqxbuTJjEwRUREpbZixQosWrRI1Y5SASlmSJELBaYCOzaFPScXRw4ewdSpUzF69GiYTCa9m+YUevfujQULFmD1qRR0q+6nlhcRFW119IWMy+uuuw7OhEGpKiQZUhKQCr6yNcxBLKpJVBY5ialI+GuH2o4YlCIiotLYvn075s6dC6O3FSG9OsDk46V3k4jKRAItQV2aw5aeia1bt6phqPfffz8DMACCgoLQrVs3/PnnnziQkInGIcwiIyrKmdRs7EvIRIsWLRAVFQVnwqCUDiQgJQU2iYiIiKhynDx5Eu+//z7sBiD0qvYwB/jq3SSicjEYjQj+vzaIXbYey5cvR40aNZwu00EvshwkKLX0eBIaBXsxWEdUhGUnktS9M+47nGcgIRERERFRBUhPT8eUKVOQkZGBoCtbwxoRrHeTiC6L0WJG6MVsvy+++AL79+/Xu0lOoUGDBujUqRMOJ2Vhe0y63s0hckqHEzOx9Xw6GjVqhPbt28PZMChFRERERG41097s2bPVrGV+zevDp251vZtEVCFMvt4qY0pm0JIswKSkC5kPnu6uu+6CxWzGj0cSkZVr17s5RE7FZrdjweEE9f97773XKbMJGZQiIiIiIrcqbL527VpYI0MQ0K6R3s0hqlDWaqFqVr74+Hg1OY8EYT1dtWrV0K9/fzWr2MqTnOWcyNGGs2k4kZKNnj17omHDhnBGDEoRERERkVs4ffq0Gtpk9LKojBKpxUPkbvya14NX7Qjs2LFD1Zgi4KabbkJISAiWn0jGiRTOdE4kYjNysPBIInx8vPHf//4XzopHaiIiIiJyebm5ufjwww+RlZWFwC4t1FAnIredka9rSzWr5Jfzv1RDVT2dj48PHnnkEeTa7fh8bxyycm16N4lI92F7X+6LQ0aODQ888KAK2jorBqWIiIiIyOX9/PPPOHjwIHwa1GQdKXJ7UvA8qEsLZGdlY+bMmbDZGIRp2bIlBgwYgHPpOVh0JFHv5hDp6rcTyWoCgO7du6NHjx5wZgxKEREREZFLO3v2LP73v/+pE/XATs30bg5RlfCuUw0+9WuqYOxvv/2md3Ocwu233446dergz9OpnI2PPHq2vaXHkxAWFoYHHngAzo5BKSIiIiJyWVLoec6cOcjOzkZAp2YwWi16N4moygR0bKpqqH399deIi4uDp7NYLHjsscfg7eWFL/fH4VQq60uRZ4nPyMEne+JgMBgxYsQI+Pn5wdkxKEVERERELktm2vvnn39U4WfJHCHyJCZvKwI6NEFGRgbmzZund3OcQq1atTBi5Ehk24DZu+OQkp2rd5OIqkRmrg2zd8eqdf7+Bx5A06ZN4QoYlCIiIiIilyQn4l9++SUMJiMCOzVXBaCJPI1Pg1qwRIZgw4YN2LVrl97NcQodOnTAbbfdhriMHMzdE4tcm13vJhFVetbwV/vjEZ2ajWuuuQa9e/eGq2BQioiIiIhc0qJFixAfHw+/FvVh9vfRuzlE+s3Gd7GWmmRLyUyUBNx0003o2rUrDiVm4asD8Wo2MiJ3tfhoIrbFpKNZs2a455574EoYlCIiIiIilxMbG6tm3DP5ecO/RQO9m0OkK0toIHwb1caJEyewatUqvZvjNMG64cOHo3Hjxth0Lg2LOSMfualV0clYeTIFNWvWxJNPPgmz2QxXwqAUEREREbmc77///kJx87aNYTCb9G4Oke782zRS24LMRJmZmal3c5yC1WrFU089hdq1auH36BSsPJmsd5OIKtSmc2lYeDgRoaEhGDduHAICAuBqGJQiIiIiIpci2SCrV6+GOSQA3vVr6N0cIqdg8vGCX/N6SEhIwJIlS/RujtPw9/fHM88+i7DQUCw6kogNZ1P1bhJRhdgTl4H5++Pg5+uLZ58dh7CwMLgiBqWIiIiIyOWypKSoa2D7xixuTuTAr3l9GL2tWLx4MdLS0vRujtOQk/Vnx42Dv5+fKga99TyXDbm2AwkZ+GRPLMxmC8Y+/TRq164NV8WgFBERERG5jOPHj2Pjxo2wRATDWiNc7+YQORWjxawK/6enp2PZsmV6N8ep1KpVC+Oeew4+Pj74fF8cdsSk690konI5nJiJWbtjAaMJY556StVNc2UMShERERGRy1i4cKG6D2jdkFlSRIXwbRQFo5cVP//yC7OlCqhfv77KmLJavfDZ3jjsjmNgilzLseQsfLQrFrkw4slRo9CqVSu4OgaliIiIiMglnDt3DuvXr4clLIhZUkTFZUs1r4e01FT88ccfejfH6VxxxRV4+plnYLJY8MmeOOyJz9C7SUSlcjw5CzP/iUG2HXj88cfRrl07uAMGpYiIiIjIJSxfvlzVkvJrVpdZUkTF8G1UGwaTSQ3hs9lsejfH6TRt2hRjx46F0WTGnN2xqmA0XR5/ixEBF2/Gi7tnuZef5Tm6/IDUh//EINMGjBw5Ep06dYK74NpBRERERE4vIyMDv//+u5phzLtOdb2bQ+TUZPieT4MaKrtw+/btejfHKbVo0UIViJbA1Ow9sRzKd5me6VAdr3atqW41/SzqMbmXn+U5Kr+jSZmY4RCQ6tq1K9wJg1JERERE5PS2bNmi6uP4qAwQdmGJSuLbuI665xC+4gNTaiif+cJQvl2xDEyR8wWkPvwnFtm2C0P23C0gJXhEJyIiIiKnt27dOnXvU6+G3k0hcgmW0ECYg/yxdds2lWlIhWvevDmeUTWmrCowtZOBKXKiWfY+/CcWOXbgsccfR+fOneGOGJQiIiIiIqcmGVLbtm2DOSRAnWQTUel416uO7KwslWlIRWvWrJkKTFmsVszdE4tt5zlrIenrQEIGZu6KQQ4MeOLJJ902ICXMejeAiMjTnU3L0bsJRC6H241n2bdvH3JycuAfFal3U8jByy+/XPjjE1+v8rZQ4byjqiFl+0H8888/6N69u97Ncfri5+Oeew5vvjkR8/bFwWYH2kf66t0s8kD74jMwe3csYDRh9KhRbjPLXlEYlCIi0klAQACsFgu+2Bend1OIXJJsP7Idkfs7cOCAurdGhOjdFCKXYg72h8FiztuGqHiNGjXC88+/gDfeeAOf74tDrt2OTtX89G4WeZA9cRmYsycWBpMZY8aMQevWreHuGJQiItJJeHg4Jr37LpKTk/VuChUiOjoaM2bMwKOPPopatWrp3RwqhASkZDsi96edUFvCg/VuCjl48cUXC33c6G2t8rZQ4QwGAyzhQeqYlpqaCj8/BlhK0qBBA7zwwoXA1Pz98ci1A12rc7lR5dsVm465e+NgNFswduxYVYjfEzAoRUSkIzmh5km1c5OAVP369fVuBlGZyUmo1EhZsmSJqsl0xRVXYO7cuejYsaN63m63q6DCrFmzkJCQgCuvvBIffvihyhRwNjKtvcnPG0Yru65EZWUJ9kfW6VicP3+eQalSqlevngpMvf7aa/j6QDxsdju612A9O6o8O2LS8dneOJitVjz99NOqzpmnYKFzIiIiIjcTHx+vgkwWi0UFpXbv3o13330XISH/Dn97++238cEHH2DmzJlYv369Olnt27evU87SlZSUxOwbonIyelnztiMqvTp16mD8hAkICgzEtwcT8OepFL2bRG5qe0waPt0bqwrtP/vssx4VkBK83ERERETkZt566y1ERUWpzCiNY8afZElNmTJFZQLcdNNN6rF58+ahWrVqWLhwIe644w44i6ysLGRmZsIrjPXDiMpDC+gyKFV2tWvXxgvjx+P111/D94cSVPHznrWYMUUVZ9v5NFVY38vbG88+O84ps5UrGzOliIiIiNzMokWL1DC92267DZGRkWrmHhmmpzly5AjOnDmDPn365D0WFBSELl26YO3atYW+pwSG5KTW8VZVNXGI6DLYL9wZjTz1K+8w/vHjJyAkOBg/HE7AGmZMUQUHpLy9fTBu3HMeGZAS3DMRERERuZnDhw/n1YdatmwZHnnkETz++OP47LPP1PMSkBKSGeVIftaeK2jixIkqcKXdJBOrKsgQRC8vL9gys6vk84jcjS3rwrbD2UrLr0aNGipjKjg4CP87xMAUVXCG1Lhxqu6jp2JQioiIiMjN2Gw2tG/fXs0eJVlSw4YNw9ChQ1X9qPIaN24cEhMT824nTpxAVfH394ctI7PKPo/IndgystQ9i5xffmBKMqa0wBRrTNHl1JDSAlLjxj3n0QEpwaAUERERkRuePDVv3jzfY1I49fjx4+r/1atXV/dnz57N9xr5WXuuIMlWCgwMzHeryrouuakZeSfXRFR62XGJauhezZo19W6Ke2RMvXAhY0pqTK09k6p3k8jF/BN7YZY9BqT+xULnOshJZFSdqKy43RARlZ7MvLdv3758j+3fvx9169bNK3ouwacVK1agbdu26jGpESWz8MlQP2fTpEkTbN++HVnnE+AdFal3c4hchj3XhuyYRNSrWxfe3t56N8ctSHDv+edfwKuvvopvD8TDYgQ6RjILjUq2Jz4Dc/fGwWL1wjPPPMuA1EUMSukg4a+dejeBiIiI3NioUaPQvXt3NXzv9ttvx4YNG/Dxxx+rm1Y8/Mknn8Rrr72m6k5JkGr8+PHqZGvgwIFwxqCUyDwdw6AUURlkxSSowFTjxo31borbFT9/7rnn8Nqrr+LLffEwGwxoG+Grd7PIiR1IyMAnu2NhMpnx9NNPc5t0wKCUDoKvbAVzEKcSJSprphQDukREpdOpUycsWLBA1YF65ZVXVNBpypQpGDx4cN5rpFOcmpqq6k0lJCSgR48eWLp0qVNmU0jgLDAoCClHzyCwQ1MYTKxAQVQa6Yei1X3nzp31borbqVOnjipQ/cbrr6v6QBajAS3CfPRuFjmho0mZmLU7FnajEWPGjFHD6elfDErpQAJSlrAgvZtBREREbuz6669Xt6JItpQErOTm7MxmM/7TsycWL16MjJPn4FO38LpXRPQvW1YOMo6dUUN1mzZtqndz3FKDBg3w9DPPqNlJP90bh+Etw9EwyEvvZpETOZ2ajY93xSLXfiFDuXXr1no3yenwMhMREREROb2rrrpK3afuOQq73a53c4icXtrBk7Dn5KptR4LQVDlkGNbo0aNVFoxkw5xM4YQMdEFMeg4+/CcG6bl2DH/kEXTo0EHvJjklBqWIiIiIyCVmvZJhidnnE5B5Kkbv5hA5NVt2DlL/OQxfX1/06tVL7+a4vVatWmHkyMeQmWvHzH9icT49W+8mkc6SsnIx858YdX/fffepCUjIRYNS0dHRuPvuuxEWFgYfHx+1wW/atCnveblSNmHCBNVRkef79OmDAwcO6NpmIiIiIqp4t912m8r4SN66n9lSRMWQjEJbZhZuvPFG+Puzlm1VkLpdDz30EFKyczFjZwwSs3L1bhLpJCPHpgJSMRk5uPXWW3Httdfq3SSn5tRBqfj4eBVRtFgsWLJkCXbv3o13330XISEhea95++238cEHH2DmzJlqGmM/Pz/07dsXGRkZuradiIiIiCpW7dq18X//93/IiU9G+sGTejeHyCnlpKQjddcRBAUF8WS4il199dW44447EJ+Zi1n/xKjgBHmWHJsdn+yJxanUbLX9DRo0SO8mOT2nLnT+1ltvISoqCnPnzs17TGaP0cgVMplJ5oUXXsBNN92kHps3bx6qVauGhQsXqh0CEREREbkP6d9J1nzyln3wqhUBk6/zzRZIpBc5P0pav0vVkrrnnnuccjZNd3fDDTcgNjYWy5cvx9w9sRjWIhwmI2t6ecr29/WBeOxPyFTDze+9917Wc3P1TKlFixahY8eOKlU7MjIS7dq1w6xZs/KeP3LkCM6cOaOG7GnkikCXLl2wdu3aIt83MzMTSUlJ+W5ERERE5PyCg4MxePBgNbNY0sY9HMZH5CDjyGlVc61t27bo1q2b3s3xSBKEkBpCch67LyFTBSm4n/IMPx9LwqZzaar4/YgRI2A0OnW4xWk49VI6fPgwPvzwQzRq1AjLli3DI488gscffxyfffaZel4CUkIyoxzJz9pzhZEpOyV4pd0kG4uIiIiIXIPMJtasWTNkHD+L9MOn9G4OkdMM25NArZeXFx588EFmaOhIghEjR45U57Ebz6Vh2fFkvZtElWztmVT8diJZ1boeM2YMrFar3k1yGU4dlLLZbGjfvj3eeOMNlSU1bNgwDB06VNWPuhzjxo1DYmJi3u3EiRMV1mYiIiIiqlxysj18+HA1s1jSht3ISUrVu0lEurLbbEj4cztsWdl44IEHEB4erneTPJ4EJSQ4US0yEkuPJ2HzuTS9m0SVZH98Br47GI/AgAA8/fTTCAgI0LtJLsWpg1ISZWzevHm+x+Sq2PHjx9X/q1evru7Pnj2b7zXys/ZcYeTqQWBgYL4bEREREbmOiIgIdbFSauckrNkOey4LCpPnStlxCNnnE9C9e3c1GQA5BznPfGrsWBVAn38gHkeSMvVuElWws2nZmLs3DiaTGaMlCFlgFBe5eFBKZt7bt29fvsf279+PunXr5hU9l+DTihUr8p6X+lAyCx/HUBMRERG5N6kj2rt3b2THJSFxw269m0Oki4yT55Cy8xAiIiM5bM8J1apVC6NGjZIcT8zeHYvYjBy9m0QVJCU7Fx/vikV6jg0PDx+uakmRmwWlZONdt26dGr538OBBzJ8/Hx9//LEqGiZkh/vkk0/itddeU0XRd+7cqSrc16xZEwMHDtS7+URERERUyWSGsQYNGiD94Emk7WdJBvIsMnQ18c8dsFitGD1qlMrIIefTokULPDhkCFKzbZi9KxYZOczsdHW5Njs+3ROngoy33HKLylIkNwxKyTSKCxYswFdffYWWLVvi1VdfxZQpU9SMKxoZs/nYY4+pelPy+pSUFCxdupTTnxIRERF5SN0WuZApNTySNu5G1rl4vZtEVCVkBsr4VVthy87Bw8OG5Y0mIed09dVXo1+/fjidlo0v9sfBxhn5XNoPhxNwMDETXbt2xc0336x3c1yaUwelxPXXX68yoDIyMrBnzx5VO8CRZEu98sorarY9ec1vv/3GtDkiIiIiDxIWFqay5w0wIP6PrchJZkFh8oDC5mu2IScxBQMGDGCWhou46667VLLFP7EZWHosSe/mUDn9eSoFf51ORb26dfHwww9zyKy7B6WIiIiIiEoik+EMGTIEtowsxP++Wc1CRuSukjbtReapGDVD+Z133ql3c6iUTCYTHn/8cVUM+9cTydh6ngF0V3MwIVNlSUkR+zFPPaUmUaPLw6AUEREREbnN8BjJGslJTEX8H9s4Ix+5pdQ9R5G27zii6tTByJEjYTTylM6V+Pv746mnnlLlZr46EI9TqVl6N4lKKT4jB3P3xsJgNKlh45KlS5ePezAiIiIichuSNSJ1RrPOxCJh7U7YWbeF3Ej60dMqSyooOBhPjx0LHx8fvZtE5ZyRTybvysq1Y87uOKRm5+rdJCqB+q72xKpi9Q888ACaNGmid5PcBoNSREREROQ2JGtETvbkhCHjyGkkb9mnd5OIKkTm6Vgk/rVTBaLGPfssszRcXIcOHXDrrbeq2ds+2xuHXAbQnZZc3Pj2YDxOpmSjd+/e6NWrl95NcisMShERERGR283IN2bMGNSqXRupu48iZddhvZtEdFmyYhOR8MdWmIxGNfSrTp06ejeJKsDAgQPRsWNH7E/IxM9HE/VuDhXhj1Mp2HQuTU2odt999+ndHLfDoBQRERERuWXdlmefeQZh4eFI3rIfqXuP6d0konLJjk9G/G+bgFwbHnvsMVXUn9wns/ORRx5BzZo1sfJkCraw8LnTOZCQgUVHEhESHKxmeTWbzXo3ye0wKEVEREREbkmGN73w/PMIDglG0sY9SDtwQu8mEZVJTmIK4n7bCHt2jgpeSL00ci8yHHP06NHw8fHG1yx87nSFzT/dG6cKmz85ahSCg4P1bpJbYlCKiIiIiNyWTL3+/HPPq+m7E9ftQtqBk3o3yS0Yva15NxgMFx40GP59jC5btgSklm+ELSMLQ4cOxZVXXql3k6iSSKbUo486Fj7nzKF6k+/iE4fC5o0aNdK7SW6LQSkiIiIicvuZrp5/XgtM/cOhfBUg4oYeqHZbL3UzhwSox+Refpbn6PJkxyUh7tcNyE3PVCfEV111ld5NoioofH7LLbeowuef74uFjYXPdS1s/t3BeJxIycbVV1/NwuaVjEEpIiIiInJ7UVFRGD9+fN5QvpR/WPycnFNWTILKkLJnZmPYsGG45ppr9G4SVZFBgwap4NTe+Ez8cixJ7+Z4rD9Pp2LjuTRcccUVuP/++/VujttjUIqIiIiIPCZj6sUJL14ofr51P5I271VXxImcRUb0ecRJUfOcXIwYMYIZUh5a+LxGjRr47UQytrHweZU7lJiJBYcTEBQUhFGjRsFisejdJLfHoBQREREReVSNqZdefFEFqFJ3H0XCmu2w5+bq3SwiVe8s/vctMBsMapav7t27690k0oGvr68qfO7t7Y35qvB5tt5N8hjxmTmYuycOBoNRbYMhISF6N8kjMChFRERERB43K9+LL76IZs2aIePYGcT+tgm2TM54RfqQbL3kbQdUvbMAf3+Mf2E8OnbsqHezSEcSNH/00UcvFj6/UGybKpe2rFOyc3HvffehSZMmejfJYzAoRUREREQex9/fH88++6zKRsk+F4+YX9YhOz5Z72aRh7Fl5yBh9Tak7DyEatWr45VXXlF1bIgkMHnrrbeqwuef7Y1FLocaV2pg+JsD8TiZkq2Kmvfp00fvJnkUBqWIiIiIyCNJrRDJRpAZr3JT0hC7dB3Sj57Wu1nkIXKSUhG7ZB0yjp9FixYt8PJLL6nhpUSagQMHolOnTtifkInFRxL1bo7bWhWdgs3n09C4cWNV2NxgMOjdJI9i1rsBRERERER6FhaWoFS9evUwfcYMVWMqOzYRAe0aw2Dk9VuqHBknzyHxrx2wZeVgwIABuOOOO2AymfRuFjnh/mn48OE4feoUVkVHo6afBZ2r+endLLeyJz4Di44kqvpRUkfKbGaIpKrxSEtEREREHk+mYX/9tdfyCqDHLluPnGTOfEUVS4rqJ27cc6GgOYwYOXIkBg8ezIAUFcnHxwdjnnoK/n5++PZgAo4kZerdJLdxNi0b8/bGwWyxYMyYMQgODta7SR6JQSkiIiIiIkBNw/7qq6+iZ8+eyI5JROzPfyP98Cm9m0VuIjshBTFL1iFt7zHUrl1brWucYY9KQ4Z1PvHkk7DDgDm74xCfkaN3k1xeWrYNs3fHIj3HprLRGjRooHeTPBaDUkREREREF8k07HKCIhksVrMFCX/tQPyf22HL4rTsVP4iyqn7jiP2l7+RE5+Mvn374rXXXkNUVJTeTSMXInXH7rv/fjU7nARTMnM5I195SdF4KR5/Pj0HgwYNQrdu3fRukkfjgEkiIiIiogIkg6VRo0aYNm0aDhw4gOyzcQjs3ALeUZF6N41ciAwBTVz7D7LOxsE/wB/DHx6O9u3b690sclEyK9yJEyewfPlyfLEvDg80C4ORRbnLHCT+4VAC9iVkqiLyUlOQ9MVMKSIiIiKiQkRERGDChAm48847YcjKRfyqLYhfsx22jCy9m0ZOzm6zI3XPUcQs/ksFpDp37oy333qbASm6bPfccw9atmyJnbEZ+Plokt7NcTlrTqXir9OpqFe3Lh555BFVTJ70xW+AiIiIiKgIUoD6hhtuwJtvvqkypzKOnsb5xX8i/cgpdcWdqKDs+GTE/roeSZv2IsDPD0888YSa1YtFlKkiyOxwsk7VrFkTK04mY/2ZVL2b5DJ2x6VjweEEhAQH46mxY9VwbdIfg1JEREREbk4CKgaDQZ0YazIyMjBixAiEhYXB399fDWE4e/asru10ZnIC+OKLL+Lee++F2QYk/LkDccs3quLVRMKWnaMCUTE//43s8wno0aMH3nnnHXTp0kXvppGb8fPzw9ixY9W+W2bkO5CQoXeTnN6p1Gx8tjceFplp76mnEBoaqneT6CLWlNJBTiKj2URlxe2GiKh8Nm7ciI8++gitW7fO9/ioUaPw888/47vvvkNQUJAq7H3zzTfjr7/+0q2tzk6GeVx33XXo0KEDPv/8c2zatAkxP/8Fv2b14N+qIYwWdq09kWTMZRw9g+Qte5Gblolq1avjgfvvv2SbI6roGflGjx6NN954HXP3xOGJNhGo5mvRu1lOKTEzFx/vikGWzY4nHhvBmfacDI+cVSggIAAWq1XN4kJEZSfbj2xHRERUOikpKRg8eDBmzZqlZvvSJCYmYs6cOZg/fz569eqlHps7dy6aNWuGdevWoWvXrjq22jVqTcnJ4NatW/HZZ5/h3K4jyDhyGgHtm8C7XnWVlUaeM1QvadMeZJ2Jg8VqwS23344BAwaobAyiyta0aVM8/PBwTJ8+HR/visWoNhHwt5r0bpZTkVkKZ+2OQUJmLu666y5V342cC4NSVSg8PBzvTpqE5ORkvZtChYiOjsaMGTPw6KOPolatWno3hwohASnZjoiIqHRkeJ6cIMuMTY5Bqc2bNyM7O1s97nhyU6dOHaxdu5ZBqVJq166dmqZ98eLF+PHHH5Hw53ZY9h1DYIemsEawfpA7y03PRMr2g0g7eFJSpVT2nAztlIAlUVW68sor1dDr77//HrN3x+LRVhGwmhgYFza7HZ/vjcPJlGz07t1bHQ/J+TAoVcXkhJon1c5NAlL169fXuxlERESX5euvv8aWLVvU8L2Czpw5A6vVeknhZRkOIs8VJjMzU900SUmc9UnIcpR6XD179sRXX32lMs1il66DT/2aCGjXGCY/FtJ1J/ZcG1L3HkPqzkOqhlTt2rXVbGitWrXSu2nkwQYNGqQCU2vWrMH8/XG4t2kojMzYxMLDifgnLkMNpb3//vuZxeqkGJQiIiIicjMnTpxQszMtX768wmYXmjhxIl5++eUKeS93JBkyjz/+uKo5JfWmDh06hIzjZ+HXvB78mteH0cput8vXjTp2BslbDyA3JQ3+AQG4/Z7bcPXVV6sZGon0JMGWoUOHIjY2Ftt270bo0UTcWN+zszX/iE7G6lMpqFMnSu2buZ06L86+R0RERORmZHjeuXPn0L59ezV9uNz++OMPfPDBB+r/khGVlZWFhISEfL8nV9qrV69e6HuOGzdO1aLSbhL4oks1btxYBe+kHEBIUBBSdh7C+R9Xq+waybIh15N5JhaxS9YhYc12ICNLDQGaMnmyGv7KE11yFrJvlwksZNTHypMp+POUfjODVvMxo7a/Rd3rYUdMusqSCgkOxtixT8PX11eXdlDp8JINERERkZuR2hk7d+7M99gDDzyg6kY988wziIqKUoWYV6xYoYaeiX379uH48ePo1q1boe/p5eWlblS6Wfp69OihCuouXbpU1ZtK2rhHBaYC2jaCd10WQ3eVIubJW/cjM/p8Xu2e22+/nXWjyGn5+fnh6aefxoTx4/G/QwkI8TahRahPlbfjnqZh0Mux5Cx8vi9OHa/GPv00wsL0awuVDoNSRERERG44MUTLli0vOVmRzrn2+JAhQ9QMcqGhoQgMDMRjjz2mAlIscl6x9aZuvPFGNcRr4cKF+HX5cpVtY9l9VNWb8qrBkyVnlJOSjpQdB5F++JQqYi7bjMzaVa9ePb2bRlQiCZpKMOaVV17BZ3vj8FirCEQFWOEJYtJzMGtXLHJhwOgnn+Q26yIYlCIiIiLyQJMnT1YZPZIpJQXM+/btq2ahpcoJEkoxbFnG3377Lf7++2/E/bYR1hphCGzXGJawIL2bSDJTV0YWUv45hLR9J2C32VC3bl3ceeedqkgykStp0KCBqqP07rvvYtbuWDzZJgKh3u596p+abcPHu2KQkp2r6mu1adNG7yZRKbn3mklEREREyqpVq/L9LAXQp0+frm5UNSIjIzFy5Ehcf/31+Oabb7B9+3bEnF6rhvPJsD5zoJ/eTfRIMote6p5jSNt9RP1fvicZpidZgxK4JXJFUlNQZpybO3cuPt4Vi8fbRMDX7J7rc7bNjjm7Y3AuPQc33XSTyk4l18GgFBERERFRFZIhJVLba8+ePZg/f/6FmfpOnIXvFbXh36ohTL4VM2MiFU8Kz6cdPIGUHYdhy8hUw1hvHnwzevXqpYpGE7m6a665Rk168fPPP2Pu7lg83DIcZqN71bOz2e2Yvz8Oh5Oy0L17dxVQJtfCvS0RERERkQ6aNWum6r5s2rRJZU6d2n9C1THybVoX/i3qw2i16N1Et2S325Fx9DSStx1Ebkqayhq8/tZb0b9/f/V/InciQ1DPnz+PDRs24NuD8bizUYhbTbSw5FgStp5PVxN5PPzww271t3kKBqWIiIiIiHQiJ1CdOnVSQ23WrFmD77//HnH/HEb6gRPwa9kAfk3qwGAy6d1MtwlGZZ2OQfLWA8iOS1LZUP369VPDfSRLisgdyRDURx99FLGxsdhw6BAivM24po57rO/rz6Ri+Ylk1KheXU3cIbPKkuthUIqIiIiISGcmkwlXXXWVGn7y66+/YuGPPyJ58z6k7T0G/7aN4FO/JjMALkN2bCKStuxD1pk4tRx79uypivzLTGVEnjAT6FNPPYUJ48fj52MxCPM2o32kL1zZ/oQMfHMwHgH+/nj6mWfg7++vd5OonBiUIiIiIiJyopNHKYQuhXoXL16MJUuWIPGvnUjdfRSBHZrAq0a43k10KTkp6UjZdgDpR06pn9u1a4c77rgDUVFRejeNqEoFBQWp4M2LEyZg/oF4hHqbUC/QC67obFo25u6Jg8lkxugxY1CtWjW9m0SXwT3L7xMRERERuTA/Pz8VPHnvvfdUVk9uQgriftuEuBWbkB2frHfznJ4tMxtJm/ci5sc1KiBVv0EDvPDCCxg7diwDUuSxatWqhSdHjYINBszZE4f4jBy4mtRsG2bvjkV6jg3DHn4YTZo00btJdJmYKUVERERE5KTCwsIwfPhwVYRbZurbsWMHMk/HwveKWvBv2xgmb6veTXQqdpsNaQdOImX7ARWYkuF5//3vf9G1a1dVW4fI07Vs2RL3338/PvnkE8zaHYsn2kTAy+Qa20auzY5P98TifHoOBg0ahCuvvFLvJlEFYFCKiIiIiMjJ1alTB88++yx27tyJzz//HCcPnETG0TPwa90Qfk3qwuAiJ5WVKfN0DJI27UVOQgp8fHxw8+Dbce2117L4MVEBffr0QXR0NJYtW4bP98bhweZhMDp5zTqZqOB/hxJwIDETnTt3VjXhyD0wKEVERERE5CJatWqFiRMnYuXKlfjuu+8uFEPffwKBHZvCu3YkPFFOcpoKRmWePKeKmPfu3Ru33XYbZ9QjKsbdd9+N06dPq+zLpceS0L9eEJzZ32dS1a1+/fp45JFHmPnoRhiUIiIiIiJysZn6rrnmGnTr1g0LFixQ2Q7xv2+BV+1IBHVuBpOfDzyBPTcXKf8cQequw7Dn2tCiRQvcc889KquMiErej4wcORLjX3gBv544h9r+VrQOd859x+HETPxwKFEFmkePHg0vL9cs0E6FY1CKiIiIiMgFyRToEoTp1asXPv30U+zatQvnz8TCX4b0NasHgxtnEmSeikHiht3ITU5DSGgo7rv3XnTq1EllShFR6fchY556CuPHj8eX++MR6WNGdT/nGu6amJmLT/fGAQYDnnjiCVVnj9yL+x6piIiIiIg8ZEat5557DiNGjECAnx+St+xHzE9/I+tcPNxNbnom4tdsU7MQ2lMzMGDAALw7aZKqMcOAFFHZ1a5dWw2Hy8y1Yc6eWGTk2OAscmx2fLInFklZubj7nnvQrFkzvZtElYBBKSIiIiIiFycBGZmJ6t1J76qhfblJqYhdth5JG/fAnpMLVydFjtOPnEbM4j9VgfdGjRrhjTfewODBg+Ht7a1384hcmgR1b7zxRjWr3TcH4tX25gwWH03EseQs/N///Z+atIDcE4fvERERERG5CT8/PzzwwAPqJG7mzJk4tfcYMqLPI7h7K1gjQ+Cq2VFJG3Yj4/hZVUvmrgceUMXMWeiYqOLI5AAHDhzA1j170PB0KnrU9Ne1PTti0vFHdApq16qFBx98kJmQbox7ciIiIiIiN3PFFVeoTKIbbrgBtpT0C1lTm/eqguCuJP34GcQs/ksFpJo3b4633npLZYIxIEVUOYXPAwMCsOBIIk6kZOnWltiMHHx1IB5WqxWPP/EEC5u7Oe7NiYiIiIjckJzQ3XnnnXjppZdQs2ZNpO4+itil65CTnAZXmFkvcf1uJPyxDWY7VPaX1M2KjIzUu2lEbiskJAQjRo6EzQ58tjdO1Zmqarl2O+btjUN6jg1DhgxRNa/IvTEoRURERETkxqT+0uuvv46rr74a2XFJiP35b6QfPQ1nlZOUipgl65C2/zjq1KmDiRMnMjuKqIq0atVKZVjGpOfgx8OJVf75K04kqzpSPXr0UMOQyf251J79zTffVGNJn3zyybzHMjIy1EwjMjWkTGl5yy234OzZs7q2k4iIiIjImcjwl6FDh6p+s8VoQsKa7SoTyW5zruF86cfOIObnv5ETn6wCUa+88gpq1Kihd7OIPMqtt96qAsJ/n0nF7rj0Kvvc48lZWHo8SZ3b33///VX2uaQvlwlKbdy4ER999BFat26d7/FRo0Zh8eLF+O677/DHH3/g1KlTuPnmm3VrJxERERGRs5IZ+qTWVN26dVUmUtyKTbBlXl7tGHOgH8yhgeq+vGS2r+QdB5Gwehu8LFY88cQTasieDEEkoqplNptVfSmL2YyvDiQgJbvyZ/DMyrXji/1xsMOARx99FL6+vpX+meQcXCIolZKSoqZ7nTVrlhrnqklMTMScOXPw3nvvoVevXujQoQPmzp2Lv//+G+vWrdO1zUREREREzkgyj6TOlEwDn3UmTg2VkyFz5RXyf20QMaC7ui9v/aiEP3cgZftBRERG4pWXX0aXLl3K3R4iunxSy+m/d9yB5KzcKhnGt+x4Es6l5aB///5o1qxZpX8eOQ+XCEpJmvGAAQPQp0+ffI9v3rwZ2dnZ+R5v2rSpSjVcu3atDi0lIiIiInKN4XyPP/44Bg4ciNzkNMQuWYesc/FV3g5bZjZil29ExtHTaNKkCV595RUWNiZyEtdddx0aNmyIjefSsC8+o9I+51RqFn6PTka1yEjcdtttlfY55JzMcHJff/01tmzZoobvFXTmzBmV0hscHJzv8WrVqqnnipKZmalumqSkpApuNRGR55H9qgyhdhfR0dH57t2FzMDFqZWJSEjh8Ntvv13tF6RMhgzlC7m6Pbyqh1XJ59syshD720ZVP0qKGkvNK4vFUiWfTUSl20fIdikzX353MAFPt68Gq8lQoZ9hs9vxzYEENePfg0OGcMiuB3LqoNSJEyfUePLly5fD29u7wt5XZvB4+eWXK+z9iIgIKiD1/PPPw93MmDED7kRm4Kpfv77ezSAiJyIBIanfMnnKFMSv3IyQq9rBq2ZEpX5mbnom4iQglZCiCprfd999nF2PyAnJKCQZtSR1nJefSMKAekEV+v5/n07Nm21PZv4jz+PUQSkZnnfu3Dm0b98+77Hc3FysXr0a06ZNw7Jly5CVlYWEhIR82VIy+1716tWLfN9x48Zh9OjR+TKloqKiKvEvISJyf3KlXQIe5PzfExFRQdLffnrsWEx6dxLif996IWOqZnilZUjFLd+AnMRUdbJ71113qRm2icg5yURiUrf59+g4dKvuh1DvigkjpOXYsOR4Enx9fHD33XdXyHuS63HqoFTv3r2xc+fOfI/JLBxSN+qZZ55RgSRJ8V2xYgVuueUW9fy+fftw/PhxdOvWrcj3lWELHLpARFSxZL/KDBwiItclWQrPPvMs3nzzTTULXui1nWEJDazQz7Dn5CLu980qIHXjjTfiv//9LwNSRC7Qx7vzzjtVYsjPRxNxT9OKGeK7/HgSUrNtGHz7zQgMrNh9DbkOp86RDQgIQMuWLfPd/Pz8EBYWpv4fFBSEIUOGqKyn33//XWVWSdBKAlJdu3bVu/lERERERC5FZr2SqeAleCRD+XJS0ivsve02O+LXbEd2TCJ69uzJgBSRC5FzbCl6vvl8uhpud7liM3Kw+nQqIiMjcO2111ZIG8k1OXVQqjQmT56M66+/XmVKycFNhu398MMPejeLiIiIiMglderUSdV4krpP8b9vhi07p0LeN3nrPmSePIfWrVvjoYceYkCKyIXI9nrPPfeo/y8+knjZ77fkWBJybXbceeddnODAwzn18L3CrFq1Kt/PUgB9+vTp6kZERERERJdPMhdkNuulS5ciaeMeBHe/vALEGSfOIXX3UdSqVUtNZGQ2u9xpCJHHa9y4Mdq1a4etW7fiUGImGgaVryTO+fRsbD6Xhnr16qFz584V3k5yLS6fKUVERERERBVPCpA3aNAA6YeikXb4VLnfJzc1HYl/74TFalEBKR8fnwptJxFVnUGDBqn7X48nlfs9lp9Ihv3iezFjkhiUIiIiIqL/b+9OYKMquwaOH0pbKG2hlKVQFqUgCGW1iGzvixD4IBEoSwIRI6tKZAcFRFRQA6KIRsOSkCigQkC0iEhEZKnmY/lYFAQpRUVlKS2+0NKVaWnvl/PAzNsCBVrg3un0/0vqtHeZeYrM5cy55zkPcAOtZpo0aZJJIqX/368muVRSlmVJ2q4jUpCbJ6NGjpL69evfl7ECsEeTJk3MFNzENJf8le4qVS+pA+ezzaJlMTEx92WMKFtISgEAAAC4qdq1a5uFhLTx+aX9CSU+P+dkkuSmXDR9qrp163ZfxgjAXgMGDDCP8WczS3zuj2czpcASiY2NFT8/0hEgKQUAAADgFrp06SLR0dHiOn3e9Ia6UwWuXMk4mGiWkx8+fDjTdAAf0axZM9MP6pcLOZLmyr/j81z5BbLvfLZUr16dXlLwICkFAAAAoFiaTNJqqYr+/pJ+IEGsgoI7Oi/jlz9MYmrIkCFSo0aN+z5OAPZdE3QxBK142pN859VS2tw850qB9OzZk8UO4EFSCgAAAMAtRUZGyv/06iX5mTmm8fnt5GdflpzfTpvpf/rhFYBv6dy5s4QEB8ue5GzJ1+zUHfjfc1lSsWJF6dGjx30fH8oOklIAAAAAbqt///4SEBgomUdOipV/62qpzKNXjxk0aJD5EArAtwQGBkrXf/1L0nPzJTHt8m2PT8rKlaSsPGnfvr1Uq1bNljGibCApBQAAAOC29IOkqZbKypGcv88Ve5xO2cv5/YxERESYflQAfJP7/f3TP7dfmfPg+avHcE3A9UhKAQAA+Ji33nrLrHYWGhpqpk/pSkmJiYlFjrl8+bKMHz/e9PoJCQmRwYMHS0pKimNjRtnQu3dv008m+8TpYo/J/uOsqZLq06cPVVKAD4uKijLJZ214nnuL6skCy5Kf/smWKlWqSJs2bWwdI7wfSSkAAAAf88MPP5iE0969e+X777+XvLw809cnKyvLc8zUqVNl06ZNsn79enN8UlKSmWoF3ErNmjWlbdu2kvdPmuSlZtyw37Isk7DSaX5du3Z1ZIwA7KEJaq18ys235NjF4qfw/Z2RK6mufHnsscckICDA1jHC+5GUAgAA8DFbtmyRkSNHSnR0tLkrvXLlSjl16pQcPHjQ7L906ZJ89NFH8t5775mGszExMbJixQrZvXu3SWQBt+JuUpxz8saG53n/SZP8jGzp1LGjBAcHOzA6AHbSHlHq11skpdwJK/exQGEkpQAAAHycJqFUeHi4edTklFZP6bLcbg8//LA0bNhQ9uzZc9PncLlckp6eXuQL5ZMmOoOCguTyqRRTGVWYblOdOnVyaHQA7PTAAw9I9ephkpB62UzTKy4ppRVSLVq0sH188H4kpQAAAHxYQUGBTJkyxUyxaNmypdmWnJxsVk4KCwsrcqz2BtF9xfWp0kbX7q8GDRrYMn54H39/f2nXrp3kZ+bIlbRMz3ZNUGlSShNWWqUHoHxM4Wvbtp1k5hXImcy8G/anufLlbFaeSUhVqlTJkTHCu5GUAgAA8GHaW+ro0aOydu3au3qeWbNmmYor99fp08U3uobvc0/DcZ39x7MtPz3LJKq055QmrgCUD+7m5YlpN07h++3aNhqcozj8awEAAOCjJkyYIN988438+OOPUr9+fc/2OnXqSG5urqSlpRWpltLV93Tfzegdbu5yw809DSc35aJIy6hr36eaR3dFHoDyQad/qz/Tc2/Yd/LatubNm9s+LpQNVEoBAAD4GJ1GpQmpDRs2yI4dO6RRo0ZF9mtjc+3vsX37ds+2xMRE0wydXkC4E1WrVjWJztzzqWIVXF0K3qUJqkIfUAGUn+tB3bp1TVLq+r5SJy+5zJRepnyjOFRKAQAA+OCUvTVr1sjGjRslNDTU0ydKe0HphwN9HDNmjEybNs00P9cPFBMnTjQJqY4dOzo9fJQRzZo1kzNnzpi+UgHhVc3Ke/p3qbhqOwC+fT2IP3dOUrKvSN3gALMtKy9fUnKuSJs20eLnRz0Mbo6/GQAAAD5m2bJlpu/T448/bu5eu7/WrVvnOeb999+Xvn37yuDBg+Xf//63SSTExcU5Om6ULe4KvLzUDCnIzTP9pHSbNj4GUL40adLEPJ7O/O8UPnfj88aNGzs2Lng/KqUAAAB8cPre7VSuXFmWLFlivoDSLgWv8i6mS8WQoCLbAJQv7vf+WU1ERVzdpqvuqQcffNDJocHLkZQCAAAAUGLu5vlXLmXKlUvB5nv6xgDlk773tUryZLpLfk9zmW0nrq28R7Iat0JSCgAAAECJ6WqM1atXl/SMbMnPyDbb6CcFlE+BgYFSr149OX3mjCw+8o9ne3BwFalZs6ajY4N3IykFAAAAoFQ0CZV6PME0O1cREdfm7QAod8aOHSuHDx++oQE6feZwKySlAAAAAJRKrVq1JCEhQXL/c8n0KQsJCXF6SAAcog3NaWqOkiIphVJzuVySlJQkvuLs2bNFHn1FZGSkKa8HAAC413T6nrJy8yQ8spbTwwEAlDEkpVBqmpCaPXu2+JqlS5eKL5k3b55nyWYAAIB7KTw8/IYEFQAAd4qkFO6qAkcTHvD+/08AAAD3Q0xMjJm+pxX0PXv2dHo4AIAyhqQUSk2nhFGBAwAAUL4rpSZNmuT0MAAAZZSf0wMAAAAAAABA+UNSCgAAAAAAALYjKQUAAAAAAADbkZQCAAAAAACA7UhKAQAAAAAAwHYkpQAAAAAAAGA7klIAAAAAAACwHUkpAAAAAAAA2I6kFAAAAAAAAGxHUgoAAAAAAAC2IykFAAAAAAAA25GUAgAAAAAAgO1ISgEAAAAAAMB2JKUAAAAAAABgO5JSAAAAAAAAsJ2//S/pfSzLMo/p6elODwUAAHgRd2zgjhXwX8RPAADgbmMoklIikpGRYR4bNGjg9FAAAICXxgrVqlVzehhehfgJAADcbQxVweLWnxQUFEhSUpKEhoZKhQoVnB4OHMzkamB9+vRpqVq1qtPDAeAwrglQGiZpMBUZGSl+fnQ9KIz4CW5cLwEUxjUBJYmhqJTSxlp+flK/fn2nhwEvoRdOLp4A3LgmgAqpmyN+wvW4XgIojGsCqt1BDMUtPwAAAAAAANiOpBQAAAAAAABsR1IKuKZSpUoyZ84c8wgAXBMA4M5wvQRQGNcElASNzgEAAAAAAGA7KqUAAAAAAABgO5JSAAAAAAAAsB1JKQAAAAAAANiOpBR8RnJyskyePFmaNGkilStXloiICOnSpYssW7ZMsrOznR4egPtg5MiRUqFCBfMVEBBg3ve9evWSjz/+WAoKCpweHgCUCcRQQPlC/ARv4u/0AIB74eTJkyZ4CgsLk/nz50urVq3Mag9HjhyR5cuXS7169aR///43nJeXl2cuxADKrj59+siKFSskPz9fUlJSZMuWLebD1RdffCFff/21+Pvf+E8d730AuIoYCiifiJ/gLaiUgk8YN26cuXAeOHBAhgwZIs2bN5eoqCiJjY2VzZs3S79+/cxxejdA7/ppcBUcHCzz5s0zF+IxY8ZIo0aNJCgoSJo1ayYffPBBkeePj4+XDh06mHM0aNPg7e+//zb7Dh8+LN27d5fQ0FCpWrWqxMTEmHEAsId+eKpTp4754PTII4/Iyy+/LBs3bpRvv/1WVq5ceVfv/StXrsikSZPM+75GjRoyc+ZMGTFihAwYMMBzjMvlMsfUrl3bVBh07dpV9u/fX+T6oa+/fft2ad++vVSpUkU6d+4siYmJNv4pAcDNEUMB5RPxE7wFSSmUeRcuXJCtW7fK+PHjzcXyZvSC5jZ37lwZOHCguQM4evRoU6Jav359Wb9+vRw7dkxee+01c1H+/PPPPRdVvYB269ZNfvnlF9mzZ48899xznud86qmnzPl6ET148KC89NJL3EEAHNajRw9p06aNxMXFlfq9r95++21ZvXq1uZO4a9cuSU9Pl6+++qrIa82YMUO+/PJLWbVqlfz0009m+kvv3r3l4sWLRY6bPXu2LFq0yHzg0g+AOgYAcBIxFIDCiJ/gCAso4/bu3WvpX+W4uLgi22vUqGEFBwebrxkzZphtetyUKVNu+5zjx4+3Bg8ebL6/cOGCOS8+Pv6mx4aGhlorV668J78LgJIZMWKEFRsbe9N9Q4cOtZo3b17q976KiIiwFi5c6Pn5ypUrVsOGDT2vmZmZaQUEBFirV6/2HJObm2tFRkZa77zzjvl5586d5vW3bdvmOWbz5s1mW05OTql+bwC4F4ihgPKJ+AnehEop+Kx9+/bJoUOHJDo62pSHumn55/WWLFliSsZr1aolISEhpofCqVOnzL7w8HDTDFAz91rCruWp586d85w7bdo0eeaZZ6Rnz56yYMEC+eOPP2z6DQHcisZShe/wl/S9f+nSJdNjQaeduFWsWNEc76bvd+2voNNR3PQuv56TkJBQ5LVat27t+b5u3brm8fz58/fs9wWAe4UYCii/iJ9gN5JSKPO01FMvnNfPL9Z+CLpP5zoXdn15+tq1a+XFF180c6O1hF2DsFGjRklubq7nGC091ZJznce8bt06adq0qezdu9dT0vrrr7/KE088ITt27JAWLVrIhg0b7uvvDOD2NKjRfgd3896/lwpPSXEHe6xwA8BJxFAArkf8BLuRlEKZp83zdAnTxYsXS1ZWVonP13nOGihpo8927dqZIOxmd+p036xZs2T37t3SsmVLWbNmjWefBlhTp041F+ZBgwaZAAyAc/TDjfY+GDx4cKnf+9WqVTNLJBduuqnNPbXvgVvjxo0lMDDQPJeb3vnTc/TDFQB4M2IoAIURP8EJJKXgE5YuXWqaaWp5qd6F0wy/3vX77LPP5Pjx46ZktDgPPfSQaZz33XffyYkTJ+TVV18tchH9888/TSCld/l0tRgNmn777TezOk1OTo5MmDDBrA6h+/TCqufqPgD20KklycnJcvbsWRPw6JLmumpU3759Zfjw4aV+76uJEyfKW2+9ZVaj0WuKLpWcmprquVOndw+ff/55mT59ullKWRt+Pvvss5KdnW3uIAKAtyOGAson4id4C3+nBwDcC5pt//nnn83FVIOfM2fOmGVONdOu5aWayS/O2LFjzblDhw41F8onn3zSHK/LoSpdflSDMl0ZQlep0bnMukqNnqdBnG7TC7fOna5Zs6a5y/f666/b+NsD5ZsGM/q+1BVZqlevblaN+fDDD83Sw35+fqV+7ytdwlgDNn2P6wczXTVKe6MU/pCmfVC0jPzpp5+WjIwM88FOAzUdCwB4O2IooHwifoK3qKDdzp0eBAAAZYEGT3oXf8iQIfLmm286PRwAAACvR/yEW6FSCgCAYrinm3Tr1s2UuWvfFZ2OMmzYMKeHBgAA4JWIn1AS9JQCAKAYWr6+cuVKefTRR82yxdr8c9u2bfQ8AQAAKAbxE0qC6XsAAAAAAACwHZVSAAAAAAAAsB1JKQAAAAAAANiOpBQAAAAAAABsR1IKAAAAAAAAtiMpBQAAAAAAANuRlAIAAAAAAIDtSEoB8BnJyckyefJkadKkiVSuXFkiIiKkS5cusmzZMsnOznZ6eAAAAF6JGAqAU/wde2UAuIdOnjxpgqewsDCZP3++tGrVSipVqiRHjhyR5cuXS7169aR///43nJeXlycBAQGOjBkAAMBpxFAAnESlFACfMG7cOPH395cDBw7IkCFDpHnz5hIVFSWxsbGyefNm6devnzmuQoUK5q6fBlfBwcEyb948s123NW7cWAIDA6VZs2by6aefep77r7/+MucdOnTIsy0tLc1si4+PNz/ro/6sr9W6dWtzl7Fjx45y9OhR2/8sAAAA7hQxFAAnkZQCUOZduHBBtm7dKuPHjzdB0s1osOM2d+5cGThwoLkDOHr0aNmwYYMpWX/hhRdMADR27FgZNWqU7Ny5s8RjmT59uixatEj2798vtWrVMoGc3kkEAADwNsRQAJxGUgpAmff777+LZVnm7lxhNWvWlJCQEPM1c+ZMz/Zhw4aZgEnvAjZs2FDeffddGTlypLlT2LRpU5k2bZoMGjTIbC+pOXPmSK9evUzp+6pVqyQlJcUEbAAAAN6GGAqA00hKAfBZ+/btM+Xi0dHR4nK5PNvbt29f5LiEhATTS6Ew/Vm3l1SnTp0834eHh5sgrzTPAwAA4BRiKAB2odE5gDJPV4rR0vLExMQi2/UungoKCiqyvbjy9OL4+V3N3+udRDfKyQEAQFlHDAXAaVRKASjzatSoYcq9Fy9eLFlZWSU+Xxt67tq1q8g2/blFixbme+1roM6dO+fZX7hhZ2F79+71fJ+amionTpwwzw8AAOBtiKEAOI1KKQA+YenSpaZcXMvKtQmnrt6id+e0Webx48clJibmlo01dbWZdu3aSc+ePWXTpk0SFxcn27Zt89wl1FVgFixYII0aNZLz58/LK6+8ctPneuONN0yAFxERIbNnzzY9GQYMGHDffm8AAIC7QQwFwFEWAPiIpKQka8KECVajRo2sgIAAKyQkxOrQoYO1cOFCKysryxyjl70NGzbccO7SpUutqKgoc17Tpk2tTz75pMj+Y8eOWZ06dbKCgoKstm3bWlu3bjXPtXPnTrNfH/XnTZs2WdHR0VZgYKB57cOHD9v02wMAAJQOMRQAp1TQ/zibFgOAsi8+Pl66d+9uys3DwsKcHg4AAECZQAwFlG/0lAIAAAAAAIDtSEoBAAAAAADAdkzfAwAAAAAAgO2olAIAAAAAAIDtSEoBAAAAAADAdiSlAAAAAAAAYDuSUgAAAAAAALAdSSkAAAAAAADYjqQUAAAAAAAAbEdSCgAAAAAAALYjKQUAAAAAAADbkZQCAAAAAACA2O3/AXW5Ay9buxmoAAAAAElFTkSuQmCC", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "import seaborn as sns\n", + "\n", + "# Create a smaller dataframe with just Dragon and Grass Pokémon\n", + "subset = df[(df[\"Type 1\"].isin([\"Dragon\",\"Grass\"])) | (df[\"Type 2\"].isin([\"Dragon\",\"Grass\"]))].copy()\n", + "subset[\"Group\"] = np.where(subset[\"Type 1\"].eq(\"Dragon\") | subset[\"Type 2\"].eq(\"Dragon\"),\n", + " \"Dragon\", \"Grass\")\n", + "\n", + "plt.figure(figsize=(12,5))\n", + "\n", + "# --- Boxplot ---\n", + "plt.subplot(1,2,1)\n", + "sns.boxplot(data=subset, x=\"Group\", y=\"HP\", palette=\"Set2\")\n", + "plt.title(\"Boxplot: HP Distribution (Dragon vs Grass)\")\n", + "\n", + "# --- Violin plot ---\n", + "plt.subplot(1,2,2)\n", + "sns.violinplot(data=subset, x=\"Group\", y=\"HP\", palette=\"Set2\", inner=\"box\")\n", + "plt.title(\"Violin Plot: HP Distribution (Dragon vs Grass)\")\n", + "\n", + "plt.tight_layout()\n", + "plt.show()\n" ] }, { @@ -337,7 +778,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 20, "metadata": {}, "outputs": [ { @@ -453,7 +894,7 @@ "4 624.0 262.0 1.9250 65500.0 " ] }, - "execution_count": 5, + "execution_count": 20, "metadata": {}, "output_type": "execute_result" } @@ -483,17 +924,91 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 21, "metadata": {}, "outputs": [], - "source": [] + "source": [ + "def euclid_dist(lon1, lat1, lon2, lat2):\n", + " return np.sqrt((lon1 - lon2)**2 + (lat1 - lat2)**2)\n", + "\n", + "# Landmarks\n", + "school = (-118.0, 34.0)\n", + "hospital = (-122.0, 37.0)\n", + "threshold = 0.50 # “close” cutoff\n" + ] }, { "cell_type": "code", - "execution_count": null, + "execution_count": 22, "metadata": {}, - "outputs": [], - "source": [] + "outputs": [ + { + "data": { + "text/plain": [ + "(6829, 10171, 246951.98213501245, 180678.44105790975)" + ] + }, + "execution_count": 22, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "PRICE_COL = \"median_house_value\" # change here if your column name differs\n", + "\n", + "# Distances to each landmark\n", + "d_school = euclid_dist(df[\"longitude\"], df[\"latitude\"], *school)\n", + "d_hospital = euclid_dist(df[\"longitude\"], df[\"latitude\"], *hospital)\n", + "\n", + "# Minimum distance to either school or hospital\n", + "d_min = np.minimum(d_school, d_hospital)\n", + "\n", + "# Boolean mask: close to either if min distance < threshold\n", + "is_close = d_min < threshold\n", + "\n", + "# Price arrays\n", + "price_close = df.loc[is_close, PRICE_COL].dropna()\n", + "price_far = df.loc[~is_close, PRICE_COL].dropna()\n", + "\n", + "len(price_close), len(price_far), price_close.mean(), price_far.mean()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "n (close, far): 6829, 10171\n", + "Mean price (close, far): 246952, 180678\n", + "Welch t = 37.992\n", + "One-tailed p (close > far) = 1.503e-301\n", + "Decision @ 5%: Reject H0 — evidence close homes are more expensive\n" + ] + } + ], + "source": [ + "# Welch t-test (independent samples, unequal variances)\n", + "t_stat, p_two = stats.ttest_ind(price_close, price_far, equal_var=False)\n", + "\n", + "# One-tailed p for alternative mean(close) > mean(far)\n", + "diff = price_close.mean() - price_far.mean()\n", + "p_one = p_two/2 if diff > 0 else 1 - p_two/2\n", + "\n", + "print(f\"n (close, far): {len(price_close)}, {len(price_far)}\")\n", + "print(f\"Mean price (close, far): {price_close.mean():.0f}, {price_far.mean():.0f}\")\n", + "print(f\"Welch t = {t_stat:.3f}\")\n", + "print(f\"One-tailed p (close > far) = {p_one:.4g}\")\n", + "\n", + "alpha = 0.05\n", + "print(\"Decision @ 5%:\",\n", + " \"Reject H0 — evidence close homes are more expensive\"\n", + " if (p_one < alpha and diff > 0) else\n", + " \"Fail to reject H0 — not enough evidence close homes are more expensive\")\n" + ] } ], "metadata": { @@ -512,7 +1027,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.10.9" + "version": "3.11.3" } }, "nbformat": 4,