Skip to content

Instantly share code, notes, and snippets.

@statkclee
Created November 30, 2015 07:06
Show Gist options
  • Save statkclee/e08b9bbf7518c4d09c4f to your computer and use it in GitHub Desktop.
Save statkclee/e08b9bbf7518c4d09c4f to your computer and use it in GitHub Desktop.
Think Stats 2, 10장 연습문제 해답
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"통계적 사고 (2판) 연습문제 ([thinkstats2.com](thinkstats2.com), [think-stat.xwmooc.org](http://think-stat.xwmooc.org))<br>\n",
"Allen Downey / 이광춘(xwMOOC)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 연습문제 10.1\n",
"\n",
"BRFSS에서 나온 데이터를 사용해서, log(체중) 대비 신장에 대한 선형 최소자승적합을 계산하라. 변수중 하나가 로그 변환된 이와 같은 모형에 대해서 추정된 모수를 나타내는 가장 좋은 방식은 어떻게 될까요? 만약 누군가의 체중을 추측하려고 한다면, 신장을 아는 것이 얼마나 도움이 될까?\n",
"\n",
"NSFG와 마찬가지로, BRFSS는 일부 집단을 과다표집(oversampling)하고 각 응답자에 대해서 표집 가중치 정보를 제공한다. BRFSS 데이터에서, 해당 가중치에 대한 변수명은 totalwt다. 가중치를 갖는, 갖지 않는 재표집을 사용해서, BRFSS에 나온 평균 응답자 신장, 평균에 대한 표준오차, 90% 신뢰구간을 추정하시오. 보정 가중치가 추정값에 얼마나 영향을 주는가?"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"BRFSS 데이터를 불러들여서, 신장과 log 체중을 추출한다."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"import brfss\n",
"import numpy as np\n",
"\n",
"%matplotlib inline"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"df = brfss.ReadBrfss(nrows=None)\n",
"df = df.dropna(subset=['htm3', 'wtkg2'])\n",
"heights, weights = df.htm3, df.wtkg2\n",
"weights = np.log10(weights)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"절편과 기울기를 추정한다."
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(0.99308041639176259, 0.0052814541694181016)"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import thinkstats2\n",
"inter, slope = thinkstats2.LeastSquares(heights, weights)\n",
"inter, slope"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"데이터에 대한 산점도와 적합선을 보여준다."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEPCAYAAACp/QjLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/Xv0Bq3auzMAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x117207d0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import thinkplot\n",
"thinkplot.Scatter(heights, weights, alpha=0.01)\n",
"fxs, fys = thinkstats2.FitLine(heights, inter, slope)\n",
"thinkplot.Plot(fxs, fys)\n",
"thinkplot.Config(xlabel='height (cm)', ylabel='log10 weight (kg)', legend=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"동일한 도식화를 하지만, 역변환을 적용해서 선형(log 아님) 척도로 체중을 나타낸다."
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEPCAYAAABcA4N7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nXAAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x135a51b0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"thinkplot.Scatter(heights, 10**weights, alpha=0.01)\n",
"fxs, fys = thinkstats2.FitLine(heights, inter, slope)\n",
"thinkplot.Plot(fxs, 10**fys)\n",
"thinkplot.Config(xlabel='height (cm)', ylabel='weight (kg)', legend=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"잔차 백분위수를 도식화한다.\n",
"\n",
"선들이 범위 대부분에 걸쳐 평평하다. 관계가 선형임을 나타낸다.\n",
"\n",
"선들이 거의 평행하다. 잔차 분산이 범위에 걸쳐 같음을 나타낸다."
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEPCAYAAACKplkeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x149090f0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"res = thinkstats2.Residuals(heights, weights, inter, slope)\n",
"df['residual'] = res\n",
"\n",
"bins = np.arange(130, 210, 5)\n",
"indices = np.digitize(df.htm3, bins)\n",
"groups = df.groupby(indices)\n",
"\n",
"means = [group.htm3.mean() for i, group in groups][1:-1]\n",
"cdfs = [thinkstats2.Cdf(group.residual) for i, group in groups][1:-1]\n",
"\n",
"thinkplot.PrePlot(3)\n",
"for percent in [75, 50, 25]:\n",
" ys = [cdf.Percentile(percent) for cdf in cdfs]\n",
" label = '%dth' % percent\n",
" thinkplot.Plot(means, ys, label=label)\n",
" \n",
"thinkplot.Config(xlabel='height (cm)', ylabel='residual weight (kg)', legend=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"상관을 계산한다."
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.53172826059835865"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"rho = thinkstats2.Corr(heights, weights)\n",
"rho"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"결정계수를 계산한다."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.28273494311894265"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"r2 = thinkstats2.CoefDetermination(weights, res)\n",
"r2"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"$R^2 = \\rho^2$ 임을 확증한다."
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"1.3378187446733136e-14"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"rho**2 - r2"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Std(ys)를 계산하는데, 신장을 사용하지 않은 예측 RMSE가 된다."
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.10320725030004974"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"std_ys = thinkstats2.Std(weights)\n",
"std_ys"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Std(res)를 계산하는데, 신장을 사용하는 예측 RMSE가 된다."
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.08740777080416139"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"std_res = thinkstats2.Std(res)\n",
"std_res"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"신장 정보가 RMSE를 얼마나 줄이는가? 약 15%"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.15308497658793585"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"1 - std_res / std_ys"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"재표본추출을 사용해서 절편과 기울기에 대한 표집분포를 계산하시오."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"t = []\n",
"for _ in range(100):\n",
" sample = thinkstats2.ResampleRows(df)\n",
" estimates = thinkstats2.LeastSquares(sample.htm3, np.log10(sample.wtkg2))\n",
" t.append(estimates)\n",
"\n",
"inters, slopes = zip(*t)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"기울기에 대한 표집분포를 도식화하시오."
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEACAYAAABcXmojAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nktRlUpAkdZkUJEld/w+djCiAjXNZhwAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x153351f0>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"<matplotlib.figure.Figure at 0x11c692b0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"cdf = thinkstats2.Cdf(slopes)\n",
"thinkplot.Cdf(cdf)\n",
"thinkplot.Show(legend=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"기울기에 대한 p-값을 계산하시오."
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.0"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pvalue = cdf[0]\n",
"pvalue"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"기울기 90% 신뢰구간을 계산하시오."
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(0.0052573684005693686, 0.0053031427019740842)"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"ci = cdf.Percentile(5), cdf.Percentile(95)\n",
"ci"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"표집분포의 평균을 계산하시오."
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.0052816475557694456"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"mean = thinkstats2.Mean(slopes)\n",
"mean"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"표집분포에 대한 표준편차를 계산하시오. 이것이 표준오차다."
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"1.3932953469804109e-05"
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stderr = thinkstats2.Std(slopes)\n",
"stderr"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"표집가중치를 사용해서 재표본추출하시오."
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def ResampleRowsWeighted(df, column='finalwt'):\n",
" \"\"\"Resamples a DataFrame using probabilities proportional to given column.\n",
"\n",
" df: DataFrame\n",
" column: string column name to use as weights\n",
"\n",
" returns: DataFrame\n",
" \"\"\"\n",
" weights = df[column]\n",
" cdf = thinkstats2.Cdf(dict(weights))\n",
" indices = cdf.Sample(len(weights))\n",
" sample = df.loc[indices]\n",
" return sample"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"표집분포를 요약하시오."
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def Summarize(estimates):\n",
" mean = thinkstats2.Mean(estimates)\n",
" stderr = thinkstats2.Std(estimates)\n",
" cdf = thinkstats2.Cdf(estimates)\n",
" ci = cdf.Percentile(5), cdf.Percentile(95)\n",
" print('mean', mean)\n",
" print('stderr', stderr)\n",
" print('ci', ci)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"가중치 없이 행을 재표본추출하고 결과를 요약하시오."
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"('mean', 168.95556271347442)\n",
"('stderr', 0.01614643033211284)\n",
"('ci', (168.93035429171971, 168.98151741142706))\n"
]
}
],
"source": [
"estimates_unweighted = [thinkstats2.ResampleRows(df).htm3.mean() for _ in range(100)]\n",
"Summarize(estimates_unweighted)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"가중치를 갖고 행을 재표본추출하시오. 만약 표집 가중치를 고려하면, 추정된 평균 신장이 거의 2cm 더 크고, 차이는 표집오차보다 훨씬 크다."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"estimates_weighted = [ResampleRowsWeighted(df).htm3.mean() for _ in range(100)]\n",
"Summarize(estimates_weighted)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 2",
"language": "python",
"name": "python2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.10"
}
},
"nbformat": 4,
"nbformat_minor": 0
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment