diff --git a/book/hte/18_hatarogeneous_treatment_effects_ko.ipynb b/book/hte/18_hatarogeneous_treatment_effects_ko.ipynb new file mode 100644 index 0000000..e7dab88 --- /dev/null +++ b/book/hte/18_hatarogeneous_treatment_effects_ko.ipynb @@ -0,0 +1,520 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "c01", + "metadata": {}, + "source": [ + "# 이질적 처치 효과 (HTE)\n", + "\n", + "타노스는 우주 인구의 절반을 무작위로 제거했다가 실패했습니다. 히어로들이 되돌려 버렸으니까요.\n", + "\n", + "타노스 주니어는 달랐습니다. 무작위 제거 대신, 제거했을 때 범죄율이 가장 크게 **증가**하는 지역을 골라 제거하겠다는 전략을 세웠습니다. 제거 후 혼란으로 범죄가 늘어난 지역이 남으면 세계가 더 빠르게 무너진다는 계산입니다.\n", + "\n", + "그런데 전략을 세우려면 먼저 이 질문에 답해야 합니다.\n", + "\n", + "**어느 지역에서 제거를 실행하면 범죄율이 가장 많이 증가하는가?**" + ] + }, + { + "cell_type": "markdown", + "id": "c02", + "metadata": {}, + "source": [ + "## 설정\n", + "\n", + "타노스 주니어는 300개 지역을 관찰합니다. 절반은 무작위로 제거 대상(T=1)으로 지정해 파일럿 실험을 합니다. 나머지 절반은 통제 지역(T=0)으로 둡니다. 제거 이후 각 지역의 범죄율 변화(crime_change)를 측정합니다.\n", + "\n", + "| 변수 | 의미 |\n", + "|------|------|\n", + "| poverty_rate | 빈곤율 |\n", + "| police_density | 경찰 밀도 |\n", + "| hero_density | 히어로 출몰 빈도 |\n", + "| population_density | 인구 밀도 |\n", + "| gang_activity | 갱 활동 지수 |\n", + "| snapped | 제거 여부 (T=1: 제거, T=0: 통제) |\n", + "| crime_change | 범죄율 변화 (양수 = 범죄 증가 = 타노스 주니어의 목표) |" + ] + }, + { + "cell_type": "markdown", + "id": "c03", + "metadata": {}, + "source": [ + "## 반사실 문제\n", + "\n", + "타노스 주니어가 각 지역의 처치 효과를 정확히 알려면, 같은 지역을 제거한 세계와 제거하지 않은 세계에서 동시에 관찰해야 합니다. 지구가 두 개여야 가능한 일입니다.\n", + "\n", + "이것이 반사실(counterfactual) 문제이고, 지역별 처치 효과(ITE)를 직접 계산하지 못하는 이유입니다.\n", + "\n", + "```\n", + "ITE_i = Y_i(1) - Y_i(0)\n", + "```\n", + "\n", + "| 지역 | 제거 여부 | Y(1) 제거 시 | Y(0) 미제거 시 | ITE |\n", + "|------|-----------|------------|--------------|-----|\n", + "| 지역 A | 제거됨 | 관찰 가능 | 관찰 불가 | 계산 불가 |\n", + "| 지역 B | 통제 | 관찰 불가 | 관찰 가능 | 계산 불가 |\n", + "\n", + "그래서 개별 ITE 대신 집단 평균인 ATE를 추정합니다.\n", + "\n", + "```\n", + "ATE = E[Y(1) - Y(0)]\n", + "```\n", + "\n", + "그런데 ATE에는 한계가 있습니다. 지역마다 처치 효과가 달라도 평균이 이를 가려버립니다. 이 이질성을 분석하는 것이 HTE이고, 특성 X를 조건으로 수치화한 것이 CATE입니다.\n", + "\n", + "| 개념 | 정의 |\n", + "|------|------|\n", + "| ITE | 지역별 처치 효과 — 직접 관측 불가 |\n", + "| ATE | ITE의 전체 평균 — 이질성 무시 |\n", + "| HTE | ITE가 지역마다 다른 현상 |\n", + "| CATE | 특성 X별 ITE 평균 — HTE를 수치화 |" + ] + }, + { + "cell_type": "code", + "execution_count": 21, + "id": "c04", + "metadata": {}, + "outputs": [], + "source": [ + "import numpy as np\n", + "import pandas as pd\n", + "import matplotlib.pyplot as plt\n", + "import matplotlib.ticker as mtick\n", + "from numpy.linalg import lstsq\n", + "from pathlib import Path\n", + "\n", + "_data = Path('thanos_junior.csv')\n", + "if not _data.exists():\n", + " _data = Path('book/hte/thanos_junior.csv')\n", + "df = pd.read_csv(_data)\n", + "\n", + "plt.rcParams['font.family'] = 'Malgun Gothic'\n", + "plt.rcParams['axes.unicode_minus'] = False\n", + "plt.rcParams['figure.dpi'] = 100\n", + "plt.rcParams['axes.spines.top'] = False\n", + "plt.rcParams['axes.spines.right'] = False" + ] + }, + { + "cell_type": "markdown", + "id": "c05", + "metadata": {}, + "source": [ + "## 1단계 — ATE: 전체 평균 처치 효과\n", + "\n", + "일단 전체 그림을 봅니다.\n", + "\n", + "```\n", + "ATE = E[Y | T=1] - E[Y | T=0]\n", + "```\n", + "\n", + "제거된 지역과 통제 지역의 평균 범죄율 변화를 단순 비교합니다. 타노스 주니어에게는 이 값이 양수(+)여야 성공입니다." + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "id": "c06", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "통제 지역 평균 범죄 변화: +1.12\n", + "제거 지역 평균 범죄 변화: -0.15\n", + "ATE: -1.27\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "mean_t0 = df.loc[df['snapped'] == 0, 'crime_change'].mean()\n", + "mean_t1 = df.loc[df['snapped'] == 1, 'crime_change'].mean()\n", + "ate = mean_t1 - mean_t0\n", + "\n", + "print(f'통제 지역 평균 범죄 변화: {mean_t0:+.2f}')\n", + "print(f'제거 지역 평균 범죄 변화: {mean_t1:+.2f}')\n", + "print(f'ATE: {ate:+.2f}')\n", + "\n", + "fig, ax = plt.subplots(figsize=(5, 4))\n", + "ax.bar(\n", + " ['통제 (T=0)', '제거 (T=1)'],\n", + " [mean_t0, mean_t1],\n", + " color=['#BBBBBB', '#4477AA'],\n", + " width=0.4,\n", + ")\n", + "ax.axhline(0, color='black', linewidth=0.8)\n", + "y_vals = [mean_t0, mean_t1]\n", + "y_range = max(y_vals) - min(y_vals)\n", + "ax.set_ylim(min(y_vals) - y_range * 0.15, max(y_vals) + y_range * 0.35)\n", + "for i, v in enumerate(y_vals):\n", + " if v >= 0:\n", + " ax.text(i, v + y_range * 0.03, f'{v:+.2f}', ha='center', va='bottom')\n", + " else:\n", + " ax.text(i, v - y_range * 0.03, f'{v:+.2f}', ha='center', va='top')\n", + "ax.set_ylabel('범죄율 변화')\n", + "ax.set_title(f'평균 처치 효과 (ATE = {ate:+.2f})')\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "c07", + "metadata": {}, + "source": [ + "## 2단계 — HTE: 효과는 모든 지역에서 같았을까?\n", + "\n", + "ATE가 음수라면 타노스 주니어의 작전은 실패처럼 보입니다. 하지만 정말 모든 지역에서 범죄가 줄었을까요?\n", + "\n", + "산점도를 보겠습니다." + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "id": "c08", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "from matplotlib.lines import Line2D\n", + "\n", + "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n", + "\n", + "ax = axes[0]\n", + "color_map = {0: '#BBBBBB', 1: '#4477AA'}\n", + "point_colors = [color_map[s] for s in df['snapped']]\n", + "ax.scatter(df['hero_density'], df['crime_change'], c=point_colors, alpha=0.5, s=20)\n", + "ax.axhline(0, color='black', linewidth=0.8, linestyle='--')\n", + "ax.set_xlabel('hero_density')\n", + "ax.set_ylabel('범죄율 변화')\n", + "ax.set_title('지역별 범죄율 변화')\n", + "legend_elements = [\n", + " Line2D([0], [0], marker='o', color='w', markerfacecolor='#BBBBBB', markersize=8, label='통제 (T=0)'),\n", + " Line2D([0], [0], marker='o', color='w', markerfacecolor='#4477AA', markersize=8, label='제거 (T=1)'),\n", + "]\n", + "ax.legend(handles=legend_elements)\n", + "\n", + "median_hero = df['hero_density'].median()\n", + "groups = {\n", + " f'히어로 저비율 (< {median_hero:.2f})': df[df['hero_density'] < median_hero],\n", + " f'히어로 고비율 (>= {median_hero:.2f})': df[df['hero_density'] >= median_hero],\n", + "}\n", + "stats = {}\n", + "for label, g in groups.items():\n", + " t0_mean = g[g['snapped'] == 0]['crime_change'].mean()\n", + " t1_mean = g[g['snapped'] == 1]['crime_change'].mean()\n", + " stats[label] = {'T0': t0_mean, 'T1': t1_mean, 'effect': t1_mean - t0_mean}\n", + "\n", + "COLORS = ['#4477AA', '#EE6677']\n", + "ax = axes[1]\n", + "labels = list(stats.keys())\n", + "effects = [stats[l]['effect'] for l in labels]\n", + "ax.bar(labels, effects, color=COLORS, width=0.4)\n", + "ax.axhline(ate, color='#888888', linestyle='--', linewidth=1.2, label=f'전체 ATE ({ate:+.2f})')\n", + "ax.axhline(0, color='black', linewidth=0.8)\n", + "y_min = min(effects)\n", + "y_max = max(effects)\n", + "ax.set_ylim(y_min - abs(y_min) * 0.2, max(y_max + abs(y_max) * 0.2, 0.4))\n", + "ax.set_ylabel('처치 효과')\n", + "ax.set_title('그룹별 처치 효과')\n", + "for i, v in enumerate(effects):\n", + " if v >= 0:\n", + " ax.text(i, v + abs(y_min) * 0.05, f'{v:+.2f}', ha='center', va='bottom')\n", + " else:\n", + " ax.text(i, v - abs(y_min) * 0.05, f'{v:+.2f}', ha='center', va='top')\n", + "ax.legend()\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "c09", + "metadata": {}, + "source": [ + "## 3단계 — CATE: HTE를 연속적으로 수치화하기\n", + "\n", + "그룹 비교로는 \"두 그룹이 다르다\"는 것만 확인했습니다. 상호작용 모델을 쓰면 특성 X에 따라 처치 효과가 어떻게 달라지는지 연속적으로 추정할 수 있습니다.\n", + "\n", + "```\n", + "Y = β₀ + β₁T + β₂X + β₃(T × X) + ε\n", + "\n", + "CATE(X) = β₁ + β₃·X\n", + "```\n", + "\n", + "β₁은 X와 무관한 기본 처치 효과이고, β₃는 X가 1 증가할 때 처치 효과가 변하는 크기입니다.\n", + "\n", + "| β₃의 부호 | 의미 |\n", + "|-----------|------|\n", + "| β₃ > 0 | X가 클수록 처치 효과 증가 |\n", + "| β₃ < 0 | X가 클수록 처치 효과 감소 |\n", + "| β₃ ≈ 0 | HTE 없음 — ATE로 충분 |\n", + "\n", + "점 추정치만이 아니라 신뢰구간도 함께 봅니다. OLS 분산-공분산 행렬에서 직접 계산합니다.\n", + "\n", + "```\n", + "SE(CATE(x)) = sqrt(Var(β₁) + x²·Var(β₃) + 2x·Cov(β₁, β₃))\n", + "```" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "id": "c10", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "beta_T (기본 처치 효과): +7.604\n", + "beta_TX (hero_density당 효과 변화): -18.245\n", + "CATE = 0 이 되는 지점: hero_density = 0.42\n", + "\n", + "hero_density=0.20 (p10): CATE=+4.00, 95% CI [+3.25, +4.75]\n", + "hero_density=0.49 (p50): CATE=-1.28, 95% CI [-1.73, -0.82]\n", + "hero_density=0.80 (p90): CATE=-7.03, 95% CI [-7.80, -6.27]\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "T_arr = df['snapped'].values\n", + "Y_arr = df['crime_change'].values\n", + "n = len(Y_arr)\n", + "\n", + "X_arr = df['hero_density'].values\n", + "X_mat = np.column_stack([np.ones(n), T_arr, X_arr, T_arr * X_arr])\n", + "coef, _, _, _ = lstsq(X_mat, Y_arr, rcond=None)\n", + "beta0, beta_T, beta_X, beta_TX = coef\n", + "\n", + "residuals = Y_arr - X_mat @ coef\n", + "sigma2 = np.sum(residuals ** 2) / (n - 4)\n", + "cov_mat = sigma2 * np.linalg.inv(X_mat.T @ X_mat)\n", + "\n", + "print(f'beta_T (기본 처치 효과): {beta_T:+.3f}')\n", + "print(f'beta_TX (hero_density당 효과 변화): {beta_TX:+.3f}')\n", + "print(f'CATE = 0 이 되는 지점: hero_density = {-beta_T / beta_TX:.2f}')\n", + "print()\n", + "for pct, label in [(0.10, 'p10'), (0.50, 'p50'), (0.90, 'p90')]:\n", + " x = float(np.quantile(X_arr, pct))\n", + " cate_val = beta_T + beta_TX * x\n", + " se = np.sqrt(cov_mat[1, 1] + x ** 2 * cov_mat[3, 3] + 2 * x * cov_mat[1, 3])\n", + " print(f'hero_density={x:.2f} ({label}): CATE={cate_val:+.2f}, 95% CI [{cate_val - 1.96 * se:+.2f}, {cate_val + 1.96 * se:+.2f}]')\n", + "\n", + "x_range = np.linspace(X_arr.min(), X_arr.max(), 200)\n", + "cate_line = beta_T + beta_TX * x_range\n", + "se_line = np.sqrt(cov_mat[1, 1] + x_range ** 2 * cov_mat[3, 3] + 2 * x_range * cov_mat[1, 3])\n", + "\n", + "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n", + "\n", + "ax = axes[0]\n", + "ax.plot(x_range, cate_line, color='#4477AA', linewidth=2, label='CATE')\n", + "ax.fill_between(\n", + " x_range,\n", + " cate_line - 1.96 * se_line,\n", + " cate_line + 1.96 * se_line,\n", + " alpha=0.2,\n", + " color='#4477AA',\n", + " label='95% CI',\n", + ")\n", + "ax.axhline(ate, color='#888888', linestyle='--', linewidth=1.2, label=f'ATE = {ate:+.2f}')\n", + "ax.axhline(0, color='black', linewidth=0.8)\n", + "ax.fill_between(x_range, cate_line, 0, where=(cate_line > 0), alpha=0.08, color='#EE6677')\n", + "ax.fill_between(x_range, cate_line, 0, where=(cate_line <= 0), alpha=0.08, color='#4477AA')\n", + "ax.set_xlabel('hero_density')\n", + "ax.set_ylabel('CATE')\n", + "ax.set_title('hero_density에 따른 처치 효과')\n", + "ax.legend()\n", + "\n", + "ax = axes[1]\n", + "cate_vals = beta_T + beta_TX * X_arr\n", + "ax.hist(cate_vals, bins=30, color='#4477AA', alpha=0.75, edgecolor='white')\n", + "ax.axvline(ate, color='#888888', linestyle='--', label=f'ATE = {ate:+.2f}')\n", + "ax.axvline(0, color='black', linewidth=0.8)\n", + "ax.set_xlabel('CATE')\n", + "ax.set_ylabel('지역 수')\n", + "ax.set_title('지역별 CATE 분포')\n", + "ax.legend()\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "c11", + "metadata": {}, + "source": [ + "### poverty_rate에 따른 처치 효과\n", + "\n", + "CATE는 어떤 변수든 조절 변수로 쓸 수 있습니다. hero_density 대신 poverty_rate로 바꿔봅니다.\n", + "\n", + "빈곤율이 높은 지역은 제거 후 자원 부족으로 범죄가 늘어날 수 있습니다. 단, 이 가설이 실제 데이터에서도 유의미한지는 신뢰구간을 함께 봐야 합니다." + ] + }, + { + "cell_type": "code", + "execution_count": 25, + "id": "c12", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "beta_T (기본 처치 효과): -4.252\n", + "beta_TX (poverty_rate당 효과 변화): +7.293\n", + "\n", + "poverty_rate=0.14 (p10): CATE=-3.20, 95% CI [-4.39, -2.00]\n", + "poverty_rate=0.37 (p50): CATE=-1.57, 95% CI [-2.31, -0.83]\n", + "poverty_rate=0.64 (p90): CATE=+0.42, 95% CI [-0.81, +1.65]\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "X_arr2 = df['poverty_rate'].values\n", + "X_mat2 = np.column_stack([np.ones(n), T_arr, X_arr2, T_arr * X_arr2])\n", + "coef2, _, _, _ = lstsq(X_mat2, Y_arr, rcond=None)\n", + "beta0_2, beta_T2, beta_X2, beta_TX2 = coef2\n", + "\n", + "residuals2 = Y_arr - X_mat2 @ coef2\n", + "sigma2_2 = np.sum(residuals2 ** 2) / (n - 4)\n", + "cov_mat2 = sigma2_2 * np.linalg.inv(X_mat2.T @ X_mat2)\n", + "\n", + "print(f'beta_T (기본 처치 효과): {beta_T2:+.3f}')\n", + "print(f'beta_TX (poverty_rate당 효과 변화): {beta_TX2:+.3f}')\n", + "print()\n", + "for pct, label in [(0.10, 'p10'), (0.50, 'p50'), (0.90, 'p90')]:\n", + " x = float(np.quantile(X_arr2, pct))\n", + " cate_val = beta_T2 + beta_TX2 * x\n", + " se = np.sqrt(cov_mat2[1, 1] + x ** 2 * cov_mat2[3, 3] + 2 * x * cov_mat2[1, 3])\n", + " print(f'poverty_rate={x:.2f} ({label}): CATE={cate_val:+.2f}, 95% CI [{cate_val - 1.96 * se:+.2f}, {cate_val + 1.96 * se:+.2f}]')\n", + "\n", + "x_range2 = np.linspace(X_arr2.min(), X_arr2.max(), 200)\n", + "cate_line2 = beta_T2 + beta_TX2 * x_range2\n", + "se_line2 = np.sqrt(cov_mat2[1, 1] + x_range2 ** 2 * cov_mat2[3, 3] + 2 * x_range2 * cov_mat2[1, 3])\n", + "\n", + "fig, ax = plt.subplots(figsize=(6, 4))\n", + "ax.plot(x_range2, cate_line2, color='#EE6677', linewidth=2, label='CATE')\n", + "ax.fill_between(\n", + " x_range2,\n", + " cate_line2 - 1.96 * se_line2,\n", + " cate_line2 + 1.96 * se_line2,\n", + " alpha=0.2,\n", + " color='#EE6677',\n", + " label='95% CI',\n", + ")\n", + "ax.axhline(ate, color='#888888', linestyle='--', linewidth=1.2, label=f'ATE = {ate:+.2f}')\n", + "ax.axhline(0, color='black', linewidth=0.8)\n", + "ax.fill_between(x_range2, cate_line2, 0, where=(cate_line2 > 0), alpha=0.08, color='#EE6677')\n", + "ax.fill_between(x_range2, cate_line2, 0, where=(cate_line2 <= 0), alpha=0.08, color='#4477AA')\n", + "ax.set_xlabel('poverty_rate')\n", + "ax.set_ylabel('CATE')\n", + "ax.set_title('poverty_rate에 따른 처치 효과')\n", + "ax.legend()\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "id": "c13", + "metadata": {}, + "source": [ + "## 왜 ATE만 보면 안 되는가\n", + "\n", + "타노스 주니어가 ATE만 봤다면 이렇게 결론 내렸을 겁니다.\n", + "\n", + "> \"제거했더니 범죄가 오히려 줄었다. 작전 실패.\"\n", + "\n", + "그런데 실제로는 이야기가 다릅니다.\n", + "\n", + "| 집단 | 처치 효과 |\n", + "|------|----------|\n", + "| 히어로 저비율 지역 | 약 +2 (범죄 증가 — 타노스 주니어의 목표) |\n", + "| 히어로 고비율 지역 | 약 -5 (범죄 크게 감소 — 최악의 타겟) |\n", + "\n", + "ATE는 -1.3이지만, 히어로 저비율 지역에서는 범죄가 오히려 늘어납니다. 처음부터 히어로 저비율 지역만 골라 제거했더라면 목표를 달성할 수 있었습니다.\n", + "\n", + "poverty_rate도 조절 변수 후보였지만, 빈곤율 상위 10% 지역의 신뢰구간이 0을 포함해 (+0.42, CI [-0.81, +1.65]) 통계적으로 불확실합니다. hero_density가 훨씬 선명한 조절 변수입니다.\n", + "\n", + "이것이 HTE가 중요한 이유입니다. ATE는 \"평균적으로 효과 있는가\"를 알려주지만, 누구에게 효과가 있는지는 알려주지 않습니다.\n", + "\n", + "---\n", + "\n", + "| 개념 | 핵심 질문 |\n", + "|------|----------|\n", + "| ATE | 처치가 전반적으로 효과 있는가? |\n", + "| HTE | 처치 효과가 지역마다 다른가? |\n", + "| CATE | 특성 X를 가진 지역에 얼마나 효과적인가? |" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": ".venv", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.13.5" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/book/hte/thanos_junior.csv b/book/hte/thanos_junior.csv new file mode 100644 index 0000000..206b1ec --- /dev/null +++ b/book/hte/thanos_junior.csv @@ -0,0 +1,301 @@ +poverty_rate,police_density,hero_density,population_density,gang_activity,snapped,crime_change +0.4944725879278468,0.401626730169318,0.8560966874441007,10.975069753055614,0.02287651008855537,1,-7.465656849642227 +0.3751527747114991,0.12399120384969056,0.34501771095886546,29.7981319307561,0.015227555890156345,1,2.174628486947467 +0.5298115439843974,0.6121491759908326,0.15809353043260516,45.102178289573345,0.24750223531406762,1,8.348629129845207 +0.2363558476234208,0.5426999757006042,0.7948426362426161,7.645621592398167,0.3914758047796728,0,1.701846744482186 +0.7680577944720322,0.5510697680674975,0.20139263162171359,47.96114185343967,0.16862281732388476,1,8.44611636141154 +0.17279156167892437,0.084529757120995,0.5101334164891866,34.54184844137954,0.3958429097499009,1,-1.8029447764622093 +0.6534810368956206,0.535500902582922,0.8143995218990053,16.396622584127837,0.2763196828530627,1,-3.1603698157081443 +0.2705978714913083,0.6607916575153117,0.13481334516761984,37.78047183659693,0.19622819417723492,0,0.8459377030069652 +0.3118692875588732,0.5577624144938228,0.6613281983618453,31.83791390863384,0.13656593869514505,1,-5.926672931752293 +0.5619785881285826,0.5587349732570068,0.14820928915303386,33.131622723487986,0.08557188823511969,0,1.3674634000487074 +0.7246439626373053,0.4867900064752157,0.5348334559185279,29.158794939779984,0.27112227384015525,0,0.6188534407403823 +0.1360938788360612,0.49546042872158513,0.028187590648553865,14.711361098701223,0.10749432564543313,0,0.24943367342870035 +0.4876026827640238,0.61954327663215,0.23848538600475147,16.645071723539754,0.18893016937852639,0,0.5001531981846119 +0.40838358575698863,0.5466008780046047,0.3150085913138461,22.96426625070618,0.15739721502150456,1,6.5825636143763635 +0.19215809699627076,0.3915047727272501,0.12625128397530963,27.424964639636492,0.08339771692831645,1,4.172348846912678 +0.5041691951652935,0.362566879568955,0.6694696290411274,39.07563383886729,0.38854617655161383,1,-4.204764442110761 +0.197928577627594,0.2972949422721982,0.5739991111366776,72.58680554108963,0.6320619557306735,0,1.359344386056544 +0.42946368034554616,0.3321451057852764,0.785668465005905,30.627532042300363,0.4093395356455052,1,-4.01128716303381 +0.26755326436227744,0.42379378974127596,0.2321616800819741,38.678241503522386,0.20099289086775948,1,5.717307049007991 +0.3207521505549691,0.6222199757514153,0.43725558246608265,43.80489260193896,0.41584856867853315,0,0.28755501336248057 +0.34709726437150484,0.28733193110655836,0.16732254569572283,40.261267331630904,0.3950104280428279,1,3.2411165207641695 +0.5551254835297625,0.34390439375227233,0.3572449649519239,14.32802773581332,0.6239912944286079,0,3.794710162840703 +0.20926312756556567,0.4471110154041621,0.8171631500926917,55.489839876807245,0.19401869088126134,0,-1.5021823097944194 +0.4466059122572963,0.3186091931981693,0.6617360135662468,50.70391994930192,0.14186127993427602,1,-3.580914829467621 +0.2981214233362209,0.3408924218496033,0.1929160629686012,28.920387126526638,0.2617493781863633,1,4.006779282847798 +0.2997950109100793,0.4045645611779177,0.06497017548977084,14.92510255319868,0.16493711980644418,1,6.349551437859647 +0.492912948873501,0.33684269090869084,0.059852513145332326,27.160091109977678,0.051775471180027295,0,2.27919378368337 +0.2815060240910244,0.7006492734737527,0.7025232506749087,8.431606505324925,0.1833232765272667,0,-0.5149360715757275 +0.29685330414082756,0.30242175505860586,0.30522247122828133,59.52235993901277,0.3779847400022724,0,0.9335137591182805 +0.40017764652283117,0.4625267036169652,0.9420742203456912,17.673212391416204,0.04914859207375694,0,0.8148978906749597 +0.17290454627902885,0.14044176105609404,0.6154502104670251,22.234146121414405,0.3399222586869775,0,1.2825133314331298 +0.3025117856919123,0.503568991052449,0.31203855668021097,57.186017803887424,0.16236477054673215,1,2.073813363086312 +0.442525056783255,0.6888868098887471,0.1979504061850249,53.42716154770821,0.3925730017549099,1,6.169886056173789 +0.07544998668702999,0.12191763331218658,0.723843540753446,32.576193009932254,0.27118220292642864,0,1.9125559970643666 +0.1616891895186843,0.47371685333299673,0.3361450248076983,64.4782074466258,0.43908155469818894,1,1.3979896886694223 +0.3095909375797926,0.6612332875282712,0.7053965185801349,14.014837650414957,0.11803429591186684,1,-4.655605306665003 +0.46937168872586077,0.6222649127172227,0.49500365282967235,32.586073139763684,0.147983052316969,0,-1.2915184920254186 +0.8128374326285992,0.21735537480006872,0.5428530280229208,14.09561824031428,0.08281769095159168,1,2.5678039450964536 +0.27137857554058564,0.3803547154072681,0.7339732748131637,48.9849176912541,0.11862712435521713,0,1.3902965819273359 +0.11444968666454716,0.4490134391136588,0.1866589187626714,48.024987086219014,0.18007437031493687,0,1.3478961533655318 +0.10835061329829057,0.5563222926437386,0.6712957337915895,67.005020765784,0.30917162896219,1,-7.084672884634046 +0.2801284540673569,0.531013138274049,0.4517178758610089,26.661544742327127,0.12590241040184189,0,-0.42738186760126184 +0.7251247646742227,0.6599300483375594,0.6986354573590743,77.29582742801732,0.25415931425689,1,-2.0620777831445105 +0.1006812023256057,0.6855705470942711,0.7053786959625228,13.291173509075541,0.4956557154554988,1,-5.680323972741218 +0.32753344786832983,0.6088617685547155,0.26698770517451464,25.796054419210456,0.35039317061284486,1,4.784768492730992 +0.4335716694002951,0.4326108816608757,0.6132239407920819,13.627141909441738,0.16341492751526698,1,-2.8528570972225618 +0.23358238496218436,0.7758293153872305,0.23861119106944292,29.999416316764997,0.004373036979611858,1,3.4021002544222534 +0.5551886133069682,0.6736026640381116,0.4381199125816716,14.46016533920854,0.26298504890290286,1,2.148126240807211 +0.6065124149754494,0.5539280363673794,0.5284459643314237,49.05980591046712,0.15206302263638594,1,0.5103537199751573 +0.14348046526929162,0.567346069801914,0.660730870948263,35.203070934771134,0.09067893525935358,1,-6.304896033655269 +0.48843173482006014,0.39023373094522895,0.6850266189405476,12.512309467547066,0.2683584936096062,0,2.7316955076374 +0.45880042986834013,0.6309326306024522,0.20999787903713343,42.752721821488535,0.4574354652024786,0,2.1487063145881304 +0.324692876431175,0.775585857995177,0.6350868790203147,61.965326476797884,0.15076507627378236,1,-0.5696454984612447 +0.22453691704495995,0.3280293562122338,0.4955881295378641,37.18264541624207,0.421751716497635,1,-2.662616965989864 +0.6355838737720279,0.34235356207957607,0.8896322115578359,10.003225582656722,0.21093650836151728,1,-4.436452936404371 +0.15151404578090366,0.6037533974523972,0.34189512872156336,24.307231081183353,0.3371732788190678,1,-0.5136942594915381 +0.257987285542402,0.34064276935452587,0.3871413315258967,39.93855017955994,0.4899046395799319,0,0.10919788594957858 +0.4802039755971562,0.4806159765105422,0.9396462230975946,22.896770408870704,0.6653806848763402,0,2.2744566144811484 +0.47492656875807726,0.4860603047013364,0.7584472169119019,13.172349671664378,0.6935419328359539,1,-3.2706585055104127 +0.6391433783830215,0.36972433297403323,0.1851608192691574,52.217550402213575,0.5567402398323105,1,12.540412137883086 +0.20753864916139775,0.807182240483515,0.40774597648804933,22.287050488450895,0.09866420917831863,1,-1.357652279983989 +0.5918839233803184,0.29650025524572565,0.6216228544655393,61.986617577646086,0.3019689222347425,0,0.3878212651376889 +0.4219542448270688,0.49254881709993936,0.7380604208641571,19.052421789021615,0.7635687923561496,1,-3.04339452825727 +0.4025995573497745,0.35421922682398643,0.31929511817284656,21.47807234857067,0.178418308975162,0,-1.9015439931438913 +0.34768973897114863,0.7426497473414765,0.8096199805128073,18.097650785875196,0.4886825959807683,0,1.1398494876271075 +0.625135189561761,0.6125014694711275,0.16510947576881183,24.052332366705564,0.2839339658269543,1,7.633650527842758 +0.1394182794470908,0.5167430988825218,0.2282176663823056,34.87006094434348,0.1896189244661487,0,1.4726501628088648 +0.10525145916987037,0.7657397337191101,0.7844631749141787,13.654113761084627,0.48490931998097586,0,1.980266657894223 +0.6784849264159828,0.4450705723029641,0.48500273798927424,11.918797356611124,0.4857760593418842,0,1.806581897267177 +0.4968065854222729,0.5744183652956263,0.5364531595624301,7.513295538834162,0.10598987156569754,1,-0.7732572474324162 +0.6404896552902578,0.3728853217380349,0.7548736072439338,6.790053301558339,0.547181187683208,1,-3.251022225066114 +0.5499458896171707,0.5024438991406982,0.3190875611412842,10.091621197742837,0.35821754169733605,0,0.1717350208939452 +0.2984813160623719,0.6122513211402678,0.17321124079598504,18.96793437938147,0.2080415429649717,0,1.164117882047743 +0.6085351045435589,0.3690209500488935,0.6247430202940095,29.675904490164044,0.4597325565000504,0,0.4656711536869844 +0.46715800479676517,0.46226439039259176,0.7326720067561643,33.75568609053764,0.19123639108671744,0,0.6997420553133438 +0.286661796356397,0.5667929291138473,0.5029299280876653,28.86472311086131,0.3001167882293854,0,-0.5993349034347416 +0.3891096460916013,0.3322691223244778,0.49295445906310686,58.07028677791314,0.43697115116674684,0,2.5335268977192094 +0.2082526245647107,0.6177968373884924,0.22961954573896093,68.40324991267047,0.34006001949308934,1,5.505590148278516 +0.9653686477138645,0.5753108962379732,0.3857585296697561,22.353393630778477,0.13113073327170685,0,0.6306745392420374 +0.0949723681550989,0.44766084715122206,0.9089509576186376,13.813419774804592,0.04642512722951789,0,-2.849421212573243 +0.0674668842496545,0.5237045240536073,0.40356910438949795,79.94853294548673,0.0474821929799521,0,-0.41366981336984077 +0.23345851531736267,0.8340832963714448,0.4372517705025126,51.21566097492825,0.21556551573187652,0,0.32096997498103375 +0.45653567190157596,0.3014285071648596,0.49135543182837266,49.10942499334909,0.09873039888511166,1,0.48844452494478563 +0.16441058869051947,0.47060848158132096,0.36858395953539536,18.476886539204674,0.011701315122322304,0,-0.6234548581166353 +0.1778739003946401,0.8678886106286428,0.7912493976537339,43.356933714587555,0.1444659588057421,0,-0.34629394007138603 +0.44693168366631825,0.35767127815464217,0.7059459464510173,29.00258268755913,0.3928455138600571,1,-2.705356517745071 +0.35175888930253063,0.41467361429158417,0.3689705074998135,25.476843005438443,0.34554575249296166,1,3.3614250983877296 +0.40594172744038687,0.6801108625586516,0.46847824508019703,25.450038547721817,0.0697023761055107,1,-3.252412130305119 +0.5781591355706356,0.5776304834389222,0.4541023657275746,35.804284677064835,0.30153981423882636,1,1.5016296235309143 +0.2837636290092098,0.5530317480396252,0.4982333815154755,59.12422720211517,0.06891221335568201,1,-1.5088223366123374 +0.3322386316273397,0.26541248849539417,0.3783491361894662,83.61560400901404,0.43484180502082104,1,0.40717019786581 +0.2501170792431812,0.6689747183975775,0.33083846285729174,20.593009144466695,0.34485701150391856,0,3.3678449899907528 +0.35280117138461503,0.8161705677252066,0.7690209053451845,29.98794122250925,0.06196258607483693,0,-0.5869567017647931 +0.3438921817873322,0.46925902258363583,0.8562843187806247,25.084880164716523,0.20066084966870953,0,0.9018001916048557 +0.4939221086201205,0.6717037917922664,0.32051321219288287,34.43849156904716,0.031198834536520288,0,1.706493348404572 +0.5471660293010695,0.414957840364608,0.44950622120917916,23.19615297529353,0.1928567794807292,1,0.9172108173384093 +0.5238483684800206,0.42519132281589683,0.6414651651971283,76.00226057884734,0.12215823984910575,1,-4.782472072052434 +0.14122055665272354,0.2624841867523439,0.16179576875442106,17.093760797165764,0.3967153517292755,0,0.3116263962895778 +0.30387624626668014,0.5595738930862458,0.7639928372482212,20.61117789022478,0.46942510379361585,1,-4.093547705676314 +0.25194098815810323,0.7424339249001304,0.3138767826004581,18.654319446328497,0.23967565398509372,0,-0.012602342633341035 +0.3571877151778369,0.4883187780385719,0.6273231971480657,24.337085649054785,0.7317855077119615,0,0.7771639307303937 +0.6804286474861757,0.665835280000734,0.0984449427504485,10.63792647407219,0.14323211128134522,0,3.050423199274566 +0.206016181181203,0.4269469138391594,0.6187447036096981,18.88621952538322,0.3433083159758156,0,0.38155591762150165 +0.24526075191077804,0.6290513760327046,0.6213535608015543,34.24265525298022,0.39337666769206825,1,-4.58455658835487 +0.35774527411740076,0.5388991256249701,0.6052277958368268,34.095513974504826,0.20198344919845362,1,-2.585140950739375 +0.2903353602842432,0.19769798637547564,0.265674160991582,31.00738958137486,0.12308700219282231,0,2.07560741329108 +0.24536581429490287,0.34075374097007965,0.3630407580592796,56.26769430832948,0.5122625676689474,1,4.929900375553727 +0.4235730492997896,0.4314262107848171,0.5226259121471785,35.03390021552531,0.2538574526337937,0,2.587862003868274 +0.3833773838280252,0.47046128346533067,0.382713990243287,55.24487768529252,0.21249972140604526,1,3.2558427390023366 +0.23146845196252155,0.5769947569234253,0.2866170856734087,83.65006663021435,0.1648452331351132,0,-0.31184823324741706 +0.09613873297388933,0.23907172906353366,0.7765279154296235,43.978577258037944,0.19829904575734245,0,-1.0928595975813953 +0.6301344047249148,0.248980299929438,0.06671305272208446,10.676181207089071,0.35350838032724285,0,2.4419366993284766 +0.24787353518414412,0.46731554234891926,0.8370830264274101,31.737986703263623,0.19717595227458815,1,-7.748841423946433 +0.7896560441926176,0.7003101625110807,0.23594636809925673,29.0357937987988,0.3453102354967876,1,6.232410968382537 +0.6965138643700532,0.5770765426221639,0.5278433140658239,61.93060934540224,0.6647253457729745,0,2.5789739471779254 +0.5111239629338078,0.3247843788297536,0.5707587121344736,24.886868069375442,0.3137363711100113,1,0.2281236139527103 +0.16946452760070668,0.18969043608343455,0.27364300387477447,15.563222389904134,0.08070974773850893,0,1.0207127347228235 +0.22515190352196196,0.39575546776885273,0.3882566960639838,19.60251582010233,0.11537195514028939,0,1.2668759331540205 +0.37881589727092807,0.5312222603026422,0.6874197578406307,10.973148017382247,0.2502382388960663,0,1.5421490045084392 +0.23153824174606305,0.7256028020322248,0.4495575621722931,1.7327671149613693,0.6656176762873836,0,0.5397342868756476 +0.3367519729312295,0.23388389503185505,0.9286236001178759,76.20300056538434,0.20075600136909355,1,-9.287679878007634 +0.06426442336872981,0.6835573266537662,0.2863368266028622,26.975162495795324,0.616931765600509,0,-1.0553276223373727 +0.6154982972956752,0.4460174869867328,0.2962854621448152,11.744748051761857,0.5675444107532533,0,5.212728785184442 +0.47281746918902084,0.7400926440965123,0.2520130107973337,65.04852722185211,0.15315029616703235,1,4.305506523515674 +0.3215326706121974,0.6081252060052632,0.34421818250559083,71.19727847849475,0.502344837218055,0,1.023882896575161 +0.8304722184404253,0.6130659602412294,0.4805182491848923,29.63272747972919,0.02252773047224856,1,1.8637228591603467 +0.25042340585642603,0.709043469533857,0.9607507492903419,10.252423447509196,0.0808289878449857,1,-6.183601474870537 +0.8285934632219212,0.4011193881005374,0.33839185791860477,19.91134507213601,0.501892265992307,1,5.35896082201214 +0.10208780360226652,0.7909113069867726,0.08067022316025248,23.253066847685254,0.10984613632359698,0,-0.9082275119418892 +0.06271301264480453,0.48252315261820367,0.40590550820006427,27.52917977756728,0.18071198255218435,1,-1.2392259668826608 +0.6083413801668794,0.6520602016010615,0.443079751467833,12.826078054736252,0.2542879494790074,0,2.5818592864740078 +0.25859017930296085,0.7511794731792999,0.8190321507927897,46.03344761452858,0.43413657566721015,0,2.329762298887529 +0.40562908937617786,0.16378699621310078,0.48842766962899603,4.943332159058052,0.07740368694240383,1,2.561396890389274 +0.5394428704266522,0.6762262375353786,0.34921397695237405,50.28880266873486,0.17130784268441832,1,3.994760179721663 +0.6768687942953532,0.2604608951030341,0.2957915961594833,56.31653053002349,0.01897704981355416,0,0.22405338746013248 +0.23217739653972053,0.18534815631937263,0.6060131170157556,21.190943186234552,0.4079016532777031,1,-3.5736020748727526 +0.3965557887272027,0.41861765212817426,0.4706481919732903,25.632215753495995,0.23065653432970984,0,1.6941113414836808 +0.3689281564344981,0.7891524737390775,0.5887801806182545,60.272269890886264,0.20144372161702676,1,0.438384188535613 +0.06314081378193165,0.45965695074726176,0.22730641986149025,39.06531889484995,0.4729549847692908,1,-0.13370343231456472 +0.44066401827662194,0.6753470042851701,0.45316333715233253,21.67503232474847,0.2257497352593228,0,2.9020092789011827 +0.5200292134820446,0.5036515669503442,0.7409467365426572,23.33060549931819,0.33715933203853055,1,-3.216239684983086 +0.4767638098130684,0.3587080215119347,0.14257850243280723,35.921877834393776,0.1282419067490359,1,7.878982839357443 +0.4330001432008056,0.8673669603789506,0.23929034748927552,72.39394593975882,0.07634856588975181,1,3.7221131304469455 +0.5295483138523914,0.6122914708302823,0.4328249604127895,23.786252248028514,0.1932352875544084,1,0.48561569633035706 +0.5422473905703746,0.810423792791967,0.7812213092003,21.255980060831195,0.4483691922105844,0,2.839365932442324 +0.2164813864622481,0.31139578843391863,0.8511060683614362,27.41198836622096,0.36419175376844,0,2.0480784191481374 +0.7652502999031201,0.23864837144641138,0.26599849482592275,27.511492758045044,0.0375163116241111,1,8.393749492672864 +0.5901646755485291,0.6143103414550057,0.7712239633733906,32.29616777096746,0.3980323063643464,0,1.9322913448513217 +0.2884704995231145,0.5386117451174078,0.8596031927698493,25.682829436022224,0.11385235308909722,0,-0.3452285753496439 +0.26028526482712555,0.48285360902503277,0.6744317614568927,51.98328547570189,0.11625211663337226,1,-7.385734191073988 +0.29002930277032835,0.6241825744952714,0.5033764576272375,14.758584374537005,0.32175599055473564,1,-0.2021232035340108 +0.5010795983421316,0.43735322003157273,0.5237333961294357,34.41515480444983,0.06280444245621516,0,1.8273286835732314 +0.6219686008921986,0.33294404700275537,0.3277873587264933,35.20823823382399,0.38305746649060246,0,0.8504027393591069 +0.7601853935975739,0.2981260067524208,0.332823100162601,15.67077770056182,0.08863008170523384,0,1.3518298275058322 +0.30990271671085856,0.5274920524662774,0.39218215169172016,34.21469445151241,0.42661136686942797,0,0.5553006134779933 +0.1515141009805998,0.3132367804505501,0.38723427760033874,35.74778284419885,0.13681467175032117,1,-2.8706741729042338 +0.2701291226646808,0.7248628546007472,0.3706284447612304,3.1577734374464783,0.21324443026249218,0,-0.714510220989421 +0.22064159056968735,0.6330357757427558,0.21372282044534482,13.801552741041828,0.3014727626387022,0,-1.0689110139658207 +0.5362351029702792,0.6631526845957019,0.8467228679099534,63.09884591423665,0.32859460289827946,0,0.9840233122797897 +0.2992993831134169,0.7523603508407097,0.4558431212630053,18.249778845851914,0.5211988860991286,1,-1.3294905608501408 +0.2628534099057459,0.5923619987716697,0.352524311525455,11.723721949020426,0.4922118247563343,1,-1.0282026488656013 +0.31152325956518545,0.3917014938963427,0.16115776510598945,6.545147836756227,0.10176862227548418,1,7.748590660687706 +0.2612207180941404,0.2699356095147892,0.6550606523289494,12.919484254152756,0.7306014524806914,0,3.853740546272861 +0.21312203048125267,0.5114016641536513,0.308160420892726,25.013050945946457,0.5597170872006924,1,0.810714076229961 +0.706804153174377,0.28917460168373404,0.37717512831387606,11.119085299740199,0.34276611062567197,1,5.377074783467647 +0.2000887994055571,0.32188948895255937,0.3190672631443641,39.361371759769355,0.16517557234405317,1,1.692247748518949 +0.5348885780150245,0.27194487596829475,0.6374359762322861,15.478574877785702,0.3075141524800508,1,0.4698618411060931 +0.12669812585070006,0.34933365795786314,0.5554087667835396,32.44253505199297,0.23643022314820453,0,-0.724286374099472 +0.4854596995999964,0.5844809014367888,0.10161201302462089,58.072313983103484,0.061713057100056395,0,2.6334844875369523 +0.7702221997736759,0.35134080228104353,0.7465704283917763,56.05209834479973,0.3490398872892305,0,1.8355476153080428 +0.22154248599301915,0.6873011911784594,0.660072510227836,32.11389584024543,0.44480389716386043,1,-6.688909813267534 +0.1913645491154335,0.6723107715440029,0.49886338197670244,12.321827694161161,0.327715150154399,0,1.6434180150828415 +0.32659683535655926,0.5393390739886259,0.2914601718445416,26.08590522776168,0.5458182710186871,1,5.1815235292396755 +0.14492342817007386,0.5928558942821915,0.7321740786799632,32.738562179152055,0.3267857559941449,1,-2.9322434847065484 +0.17815612665276545,0.4636641951339211,0.5019185592448059,92.29979584547849,0.34976875832051996,1,-3.994843502659818 +0.8878273298841941,0.562773649682847,0.37700691215211046,6.491550645837552,0.4109086964039715,1,7.511744820966037 +0.5849521848254172,0.18585441521918564,0.2097502926205175,26.142992325955934,0.11263638434749562,1,5.913780884044947 +0.17274959889390795,0.5821332443992043,0.46383286192920214,38.55394674373601,0.11408987899469489,0,1.3393333760579562 +0.6394960911919297,0.8107335062422533,0.8124803934594235,28.727442287914492,0.15074882891645208,0,-0.7964819016804361 +0.27486291973767324,0.44065769395392324,0.43161364249401313,30.651551543814932,0.26716708395246563,0,-0.9758894303838184 +0.44578159946662627,0.5476707502381177,0.16095916776200628,16.119439653961965,0.0307621369863238,1,4.464810664468327 +0.3536606382434299,0.4385511037909999,0.3913256389970091,45.57837119891681,0.19391724378471065,1,1.2068136945252528 +0.5926253815330851,0.35497450138367753,0.3894983189462793,27.519214345275035,0.40487448795550063,1,2.8158112674845293 +0.4685208666726514,0.665046480636946,0.49130242056513734,30.585084396408185,0.42513310073375654,0,2.5793638367186507 +0.22700214373887528,0.4277114566692738,0.3021315477005372,48.338626403412654,0.03145402806900397,0,1.8565449580722315 +0.1783136037865479,0.6648577765000528,0.34643731570159086,22.097817369783467,0.2601901888686549,0,1.2898297244649344 +0.12896098997411476,0.5165980762522916,0.15446815764996127,54.34113973417781,0.2568348841118217,1,5.805475223530383 +0.5199648029053999,0.2250743180004265,0.5700085038892293,10.838856756890218,0.3414606538046598,0,2.6414813425081074 +0.12429073385249477,0.323233431810297,0.30653602561224946,31.632474184466876,0.18769985681103799,1,-1.1722106400465093 +0.36739844279058403,0.5448136085381636,0.24327729951627922,15.4448141168864,0.3303751480435607,0,3.2631967862282196 +0.3626600147192707,0.5027591526473508,0.72256976869483,12.769723546098618,0.12165282713124273,0,-0.07920780763604951 +0.31795004220798584,0.40936118934918386,0.7750588133260653,71.92506070845498,0.0647414919665805,0,2.0395996707362896 +0.4587024461631462,0.5101364996810365,0.5369446838874926,11.267000530432778,0.44988621380617055,0,4.126344591253403 +0.1476335857893545,0.3552870619929971,0.8470199419028595,27.847241235862832,0.38678744704379114,1,-7.814336833957094 +0.1588352786669392,0.31602841602788095,0.7618356548008954,10.63012082377342,0.43716857917407165,0,-1.2094236414258948 +0.4923670945381023,0.5627819406612331,0.4843764990894026,52.46287979389612,0.22971222445829187,0,-0.003574457895029859 +0.23791011371752102,0.05752402716494482,0.2361844190634108,20.705480753172157,0.12169689960049644,1,5.059857952680046 +0.5094639219664361,0.3908495053217242,0.39389950243971267,48.19459426169726,0.06587006750479728,1,1.9821908550067424 +0.36685039274302544,0.5350013305386517,0.31093453858616893,37.79846579827946,0.2398559141736382,1,2.768902640358061 +0.4931802960520734,0.7744953474864632,0.6339811640891277,84.74675128499008,0.22548186825548683,1,-1.1583332758657137 +0.2863351107160408,0.465984750456171,0.3920390880569252,9.068869017406252,0.2920166839456225,1,1.6828453219752721 +0.2138427255566912,0.7997075862996245,0.1380822989258208,13.720508890076776,0.11176246780175729,0,0.48354524455215475 +0.8015673746907723,0.3392284519039431,0.43470124516772524,46.72583663817352,0.08994164170346762,1,3.917136233870191 +0.8356382413698962,0.6359762163079086,0.4481161070241339,32.54115024613078,0.13560000945201514,1,3.6208753798922424 +0.7258282511419661,0.4746737096884163,0.5059188892937314,23.12181004072936,0.30014061878164416,0,3.589940519237785 +0.4390802716014603,0.2965396628082126,0.6924507751066126,76.0991362267584,0.44769370280476534,0,2.746442854207389 +0.36851622783961213,0.7206076278741576,0.15057887998859024,26.67781860776151,0.12400645797088794,0,0.2470039706597703 +0.08942328063760928,0.4854615558541853,0.7577611085930716,15.851387058936282,0.07882283890062938,1,-6.539537837500021 +0.4339357233916055,0.25056848019462985,0.5323056538470203,48.524994965916484,0.15631339768041402,0,1.1703143126463393 +0.22086687120191514,0.6181234656533967,0.6424497718054418,27.226200447552955,0.653728890851987,0,0.07444166815473285 +0.12835577061397604,0.21958934609803718,0.3569247549149241,28.527033146199948,0.12519251749106985,1,-1.3771040124137641 +0.43108478664177624,0.647444855905698,0.444393452947237,39.6565537381456,0.29841537814050545,1,-0.6380444273636152 +0.8350789220957977,0.3175792358229684,0.7494439831178427,40.552145050859224,0.14289966672080395,0,2.064841839000502 +0.12190552820110218,0.9037586174447436,0.7350919808020957,13.590639915178171,0.16383244918226905,0,1.6402092830273907 +0.3334759171519496,0.40256300342680135,0.5103166929568481,22.290376642433465,0.4475756867443481,0,0.2853733499719806 +0.23912817572010373,0.4862922304572836,0.39042659998030804,9.9435885671599,0.6095768626128513,0,1.7329596634558844 +0.49101342630054967,0.5793489894141686,0.6004027172408981,32.82795125971074,0.5690105395875431,0,1.046678350221447 +0.511211412223011,0.37242946849735586,0.5124104674716627,71.21231214932703,0.4895798864359389,0,1.7499009791363984 +0.2671401804991139,0.7248165892732064,0.8138762898935565,13.989479982676688,0.16869494745881963,1,-9.898000105925439 +0.25930882747352907,0.536315183912257,0.29520718095650467,8.574065894903617,0.13989725237984835,1,1.4327221671573116 +0.5633483274837263,0.1750154220790777,0.3006617395606568,17.610957640845097,0.10085952617536036,1,4.394369195825914 +0.21429159425545574,0.22304821158332516,0.40314466364394286,22.995918987123865,0.36187236134500655,0,1.954003209926598 +0.12091171166032434,0.7481127562860774,0.2460788113518518,40.758684746561435,0.1060566038109172,0,0.16165695619978587 +0.4382396279697666,0.441609231373014,0.5657363669006568,21.503323968411596,0.5113876405166335,0,0.09829674614148476 +0.376975564187992,0.48156532644343925,0.12343958463646158,33.17771749791859,0.06701792235295127,1,7.449635438563398 +0.404665808769787,0.4381414786862942,0.8557606435675232,19.34710690447399,0.013073253529156472,0,1.1174570150143728 +0.606987437278647,0.38039859761244776,0.6260018191913919,23.795052396718958,0.49400419465458145,1,0.10005425752571523 +0.5685424169743103,0.483025061637512,0.6768308095192551,15.494505399828737,0.562125220822519,1,-1.4386209686479816 +0.3753230187186548,0.43339729666801446,0.12041704065050118,12.297772942321181,0.2308008538206819,0,0.3775372962959842 +0.3496972339426674,0.8823754019273647,0.6348671855518568,41.60273676363329,0.3189737236309907,0,1.1313672567857078 +0.1971549165598044,0.5821807789845124,0.5604601287976957,20.22780525460969,0.5595794479961801,0,1.8418255205801346 +0.11564662621161198,0.5387247873423501,0.575410711603741,34.60916081700451,0.28340735269553297,1,-6.908826791932447 +0.3820814096836399,0.48298043575656424,0.45808457610368974,15.3284878708822,0.2488467066671865,1,-0.6002179611549913 +0.5250642812117821,0.5146296318592494,0.6359207521355592,26.78406239225368,0.07704038908425809,1,-2.4639643168301455 +0.6442504342419522,0.48181325441458006,0.28788992277170683,5.045803219244119,0.1935248990474103,1,5.905727750508504 +0.6785352928743399,0.6102088546528616,0.5917073897565386,21.208759019178054,0.05772390497814107,0,3.330811342817129 +0.6802563009155161,0.4637435252882793,0.2172671548522801,16.456669528023266,0.5677723223768842,0,1.5549588003168644 +0.4978758056806321,0.7503243196763087,0.17800895617839957,51.27085091604114,0.018928530330750013,0,0.03421738780032468 +0.24758339801843504,0.4302055253165481,0.8681506131335458,10.948477841302457,0.680931770328757,1,-8.235603604866768 +0.3008760967845904,0.6938025601721709,0.277555820422343,24.412124145712006,0.2602705195112872,0,-0.4729246159526379 +0.5350341647961923,0.7588028944879467,0.23102135414044367,32.588582549039764,0.27661387826250616,1,4.9305682280064165 +0.4285536042063098,0.34686489153018896,0.920403115792557,47.25769007204411,0.5642803902480515,0,3.2235600315668 +0.5762687756058548,0.22588264627450969,0.7670917391304015,43.598511368901576,0.15608953394074307,1,-5.115689145334078 +0.3788974160591878,0.7872236420961554,0.38671919798082577,25.624193493408413,0.2207432200245713,1,1.7451155715601687 +0.3939769007206501,0.44594096880550427,0.968168193956265,27.87448559301883,0.44431873006364747,0,-0.9282261515437622 +0.35675244305010617,0.5226333661511523,0.3533867846125684,51.244907215363,0.4696116916848091,0,0.6448403517054755 +0.30165483367553575,0.35244304098377727,0.7658363598070336,25.32278735129146,0.25123445766608976,1,-6.414369049425359 +0.7453134388760781,0.38750735211659065,0.5813169993380412,29.692472669289586,0.25920943036015637,1,3.8873159797747827 +0.2775652102142416,0.7908233359868805,0.350560694308637,52.19746874798095,0.5908043757728975,0,0.8112775890726069 +0.47224039020888076,0.7458329462811468,0.7386885069482529,91.31809097827784,0.19020991044662997,1,-7.7135079487528895 +0.24193215672004803,0.5389837004456781,0.15219830786242938,31.228325961888647,0.20635409723670792,1,1.3132110026925665 +0.7324688482988868,0.457706063704139,0.5902096988362532,36.35903905438749,0.346115254396718,1,1.2540138802467415 +0.17518347123494732,0.6211705340212865,0.43780353797500143,42.85832749584006,0.21302143590802894,1,-0.5782838209558662 +0.21362332506949522,0.3924117972459337,0.7402168860560522,51.916381321586385,0.20207603421352077,0,1.2157940328304724 +0.5885952054469751,0.558827274444623,0.5174176698898393,47.92829884061901,0.35772085236029555,0,2.1737552053183786 +0.4552906053015979,0.3198342866930708,0.5844036644235476,18.37734821973848,0.4134414801362291,0,3.616309885742681 +0.30020339884746133,0.28964744022523164,0.6242161471201758,32.47522916914213,0.11270138219368593,0,0.026978689598826988 +0.5246563455568706,0.5583607599420901,0.3990087547245326,48.116924357127445,0.25904141415921694,1,3.8418374273868303 +0.35982282857973474,0.528360418664186,0.5632401351531274,58.11664254143398,0.48729173297670303,1,-3.9017162557890868 +0.28780111305753925,0.5060932195689483,0.9115863109552153,24.439408043923585,0.3546031162160421,1,-9.972652316219985 +0.30883596773519334,0.4049516923496323,0.5456934048646637,32.0119808480813,0.2772597045723612,1,-4.60395195354984 +0.6709110003418132,0.5970654529945577,0.3998659042333558,34.2083882268692,0.3341160507488601,1,3.499511424374389 +0.5508477402063013,0.41713430129063234,0.7360652438976234,27.316035087508528,0.07653005017681005,0,1.319077193782378 +0.4894569314968961,0.1688589761981891,0.3856525734648088,31.52378201732087,0.2059000288827377,0,1.8824641143977545 +0.40043583581665615,0.4634197522261218,0.8755658272220762,4.801635575375731,0.11896529321866238,1,-7.288926504105563 +0.3307580505369147,0.33505374809015526,0.8150883411773739,122.42123546904719,0.279307444743449,0,-0.1754947032101053 +0.264427484282743,0.6434482321849542,0.37405438526241414,23.81200672839465,0.5533396927716132,0,0.9711523395866777 +0.62942426566283,0.7757010275218063,0.3704811808738051,24.96760983462377,0.10607194385685235,0,-0.9214543457282347 +0.14265316365533373,0.841279287363319,0.6702667206362294,7.864741637047974,0.24021186689745994,0,2.111947400095978 +0.10287446661922546,0.49620131820807994,0.851535974505375,6.839486424143764,0.1374742704263033,0,-0.36806848684479837 +0.5142275729647947,0.35534005997274937,0.7208733936579604,35.631345053606545,0.10066702436786426,0,0.6553475290918264 +0.28721810145061505,0.515850153053618,0.48473342364005595,48.22833198855122,0.31560890708575234,0,3.3604324581358562 +0.6628476793886204,0.5323793961484639,0.49472398756249636,43.27520441303517,0.08530236625691542,0,-0.23660142579261567 +0.5689692399510722,0.5718971072617233,0.46578654285784205,15.299801430155092,0.330231980826057,1,0.13675157146093042 +0.37627642451334165,0.6955117072453664,0.5244108882654301,41.829441146450165,0.3520156295935354,1,-2.688719103518812 +0.5737422142149315,0.407718921355464,0.3722394970446159,24.722729435687768,0.14958925135173387,0,1.5108791514435758 +0.6177611153946437,0.5936817845188871,0.9669419600261788,21.06130474083546,0.2600317605561848,1,-6.27127606086955 +0.6452270584349333,0.3305675682247108,0.7004473803055027,10.15498003362239,0.7036789230925283,1,0.38529416055244736 +0.31784002116188664,0.4715826296311664,0.6443990439134487,16.615749722365635,0.01533030467549973,0,4.3705910194479465 +0.10092718255948747,0.8810606913301108,0.8014780363634179,36.11703427157864,0.19970783920290563,1,-12.374053705731578 +0.020521665464050796,0.3491029072151716,0.15053643691871432,32.272128362953424,0.6266366464542201,0,1.8788387657720462 +0.6230024563937469,0.6504908130362703,0.7121935499406467,29.506295373437066,0.3365087743520674,0,1.0558382459907583 +0.3763702905238614,0.4410904297596576,0.8094170358892551,4.573780072789715,0.3959378836007569,0,0.60027348568028 +0.1981868143950164,0.6590066014990927,0.21050970220918436,29.688647540563863,0.5441942903353034,1,2.4942304765805265 +0.4270412198312654,0.6692716213523916,0.7991901143121325,18.19725561611563,0.4657759948536723,1,-6.642925216994002 +0.4097754356691024,0.36158121793382864,0.6986865938509363,32.410080101875984,0.3227798867334963,0,4.040092114445649 +0.4196654890027737,0.6035477546097536,0.500117043279497,52.80355686517551,0.3859515749196846,0,2.690797852758791 +0.307862279444598,0.23531828137513175,0.4430363988822813,19.996449773835636,0.1145532275846985,0,0.9521689799641215 +0.533943598018742,0.2763143758694923,0.2634849229918877,28.391720078631092,0.2096375736432129,1,7.232248168462134 +0.17919895769844274,0.5557255839186531,0.10545453413887505,75.84525669228796,0.5530720505597887,1,2.9962340740388447 +0.32543741928116404,0.7456969859644602,0.7612077266481396,31.45717633430609,0.5649825824925555,0,-0.6354772601322463 +0.22606070242179588,0.580150575148445,0.23085740617040065,26.304504461718075,0.2784139596418545,1,-0.411161243427816 +0.5268627733425103,0.3674295342789147,0.3222592763923667,15.471621505933943,0.23350176494174021,0,0.6310405054471651 +0.23602468787738498,0.1321713473206377,0.9352026417529582,42.49663019831169,0.11477506368630316,1,-9.369498790547684 +0.13503875413068298,0.6063366241267034,0.6253538846791761,13.699682944791347,0.3788431065556675,1,-4.735383681165889 +0.45687489393324293,0.4771946505053858,0.580247954858926,54.11897287030486,0.014756644952194309,1,-1.5562576360726583 +0.4460760411009474,0.5060759427819777,0.4642692601383799,39.06059205412675,0.12381020572892627,1,-1.7172030905055593 +0.135967393823498,0.5696450815795717,0.7652095375086789,18.618239713503357,0.31721133972463256,0,1.6774118932612205 +0.6031738465403812,0.6737056993828684,0.222671159561585,15.826572564749075,0.2911707963621436,0,2.077598710321059 +0.4135981319395521,0.48167986397519114,0.6247588463724525,16.085150982876133,0.23943967549089876,1,-1.426495138042573