{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import numpy as np"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
">**作者** : Gaël Varoquaux\n",
"\n",
"**必要条件**\n",
"- 标准Python科学计算环境 (numpy, scipy, matplotlib)\n",
"- [Pandas](http://pandas.pydata.org/)\n",
"- [Statsmodels](http://statsmodels.sourceforge.net/)\n",
"- [Seaborn](http://stanford.edu/~mwaskom/software/seaborn/)\n",
"\n",
"要安装Python及这些依赖,推荐下载[Anaconda Python](http://continuum.io/downloads) 或 [Enthought Canopy](https://store.enthought.com/), 如果你使用Ubuntu或其他linux更应该使用包管理器。\n",
"\n",
"---\n",
"\n",
"**也可以看一下: Python中的贝叶斯统计**\n",
"\n",
"本章并不会涉及贝叶斯统计工具。适用于贝叶斯模型的是[PyMC](http://pymc-devs.github.io/pymc), 在Python中实现了概率编程语言。\n",
"\n",
"---\n",
"\n",
"**为什么统计学要用Python?**\n",
"\n",
"R是一门专注于统计学的语言。Python是带有统计学模块的通用编程语言。R比Python有更多的统计分析功能,以及专用的语法。但是,当面对构建复杂的分析管道,混合统计学以及例如图像分析、文本挖掘或者物理实验控制,Python的富有就是物价的优势。\n",
"\n",
"---\n",
"\n",
"**内容**\n",
"- 数据表征和交互\n",
" - 数据作为表格\n",
" - panda data-frame\n",
"- 假设检验: 对比两个组\n",
" - Student’s t-test: 最简单的统计检验\n",
" - 配对实验: 对同一个体的重复测量\n",
"- 线性模型、多因素和方差分析\n",
" - 用“公式” 来在Python中指定统计模型\n",
" - 多元回归: 包含多元素\n",
" - 事后假设检验: 方差分析 (ANOVA)\n",
"- 更多的可视化: 用seaborn来进行统计学探索\n",
" - 配对图: 散点矩阵\n",
" - lmplot: 绘制一个单变量回归\n",
"- 交互作用检验\n",
"\n",
"---\n",
"\n",
"**免责声明: 性别问题**\n",
"\n",
"本教程中的一些实例选自性别问题。其原因是在这种问题上这种控制的声明实际上影响了很多人。\n",
"\n",
"## 3.1.1 数据表征和交互\n",
"\n",
"### 3.1.1.1 数据作为表格\n",
"\n",
"统计分析中我们关注的设定是通过一组不同的属性或特征来描述多个观察或样本。然后这个数据可以被视为2D表格,或矩阵,列是数据的不同属性,行是观察。例如包含在[examples/brain_size.csv](http://www.scipy-lectures.org/_downloads/brain_size.csv)的数据:\n",
"\n",
"`\n",
"\"\";\"Gender\";\"FSIQ\";\"VIQ\";\"PIQ\";\"Weight\";\"Height\";\"MRI_Count\"\n",
"\"1\";\"Female\";133;132;124;\"118\";\"64.5\";816932\n",
"\"2\";\"Male\";140;150;124;\".\";\"72.5\";1001121\n",
"\"3\";\"Male\";139;123;150;\"143\";\"73.3\";1038437\n",
"\"4\";\"Male\";133;129;128;\"172\";\"68.8\";965353\n",
"\"5\";\"Female\";137;132;134;\"147\";\"65.0\";951545\n",
"`\n",
"\n",
"### 3.1.1.2 panda data-frame\n",
"\n",
"我们将会在来自[pandas](http://pandas.pydata.org/)模块的[pandas.DataFrame](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html#pandas.DataFrame)中存储和操作这个数据。它是电子表格程序在Python中的一个等价物。它与2D `numpy`数据的区别在于列带有名字,可以在列中存储混合的数据类型,并且有精妙的选择和透视表机制。\n",
"\n",
"#### 3.1.1.2.1 创建dataframes: 读取数据文件或转化数组\n",
"\n",
"**从CSV文件读取**: 使用上面的CSV文件,给出了大脑大小重量和IQ (Willerman et al. 1991) 的观察值 , 数据混合了数量值和类型值:"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"
\n",
" \n",
" \n",
" | \n",
" Unnamed: 0 | \n",
" Gender | \n",
" FSIQ | \n",
" VIQ | \n",
" PIQ | \n",
" Weight | \n",
" Height | \n",
" MRI_Count | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" 1 | \n",
" Female | \n",
" 133 | \n",
" 132 | \n",
" 124 | \n",
" 118 | \n",
" 64.5 | \n",
" 816932 | \n",
"
\n",
" \n",
" 1 | \n",
" 2 | \n",
" Male | \n",
" 140 | \n",
" 150 | \n",
" 124 | \n",
" NaN | \n",
" 72.5 | \n",
" 1001121 | \n",
"
\n",
" \n",
" 2 | \n",
" 3 | \n",
" Male | \n",
" 139 | \n",
" 123 | \n",
" 150 | \n",
" 143 | \n",
" 73.3 | \n",
" 1038437 | \n",
"
\n",
" \n",
" 3 | \n",
" 4 | \n",
" Male | \n",
" 133 | \n",
" 129 | \n",
" 128 | \n",
" 172 | \n",
" 68.8 | \n",
" 965353 | \n",
"
\n",
" \n",
" 4 | \n",
" 5 | \n",
" Female | \n",
" 137 | \n",
" 132 | \n",
" 134 | \n",
" 147 | \n",
" 65.0 | \n",
" 951545 | \n",
"
\n",
" \n",
" 5 | \n",
" 6 | \n",
" Female | \n",
" 99 | \n",
" 90 | \n",
" 110 | \n",
" 146 | \n",
" 69.0 | \n",
" 928799 | \n",
"
\n",
" \n",
" 6 | \n",
" 7 | \n",
" Female | \n",
" 138 | \n",
" 136 | \n",
" 131 | \n",
" 138 | \n",
" 64.5 | \n",
" 991305 | \n",
"
\n",
" \n",
" 7 | \n",
" 8 | \n",
" Female | \n",
" 92 | \n",
" 90 | \n",
" 98 | \n",
" 175 | \n",
" 66.0 | \n",
" 854258 | \n",
"
\n",
" \n",
" 8 | \n",
" 9 | \n",
" Male | \n",
" 89 | \n",
" 93 | \n",
" 84 | \n",
" 134 | \n",
" 66.3 | \n",
" 904858 | \n",
"
\n",
" \n",
" 9 | \n",
" 10 | \n",
" Male | \n",
" 133 | \n",
" 114 | \n",
" 147 | \n",
" 172 | \n",
" 68.8 | \n",
" 955466 | \n",
"
\n",
" \n",
" 10 | \n",
" 11 | \n",
" Female | \n",
" 132 | \n",
" 129 | \n",
" 124 | \n",
" 118 | \n",
" 64.5 | \n",
" 833868 | \n",
"
\n",
" \n",
" 11 | \n",
" 12 | \n",
" Male | \n",
" 141 | \n",
" 150 | \n",
" 128 | \n",
" 151 | \n",
" 70.0 | \n",
" 1079549 | \n",
"
\n",
" \n",
" 12 | \n",
" 13 | \n",
" Male | \n",
" 135 | \n",
" 129 | \n",
" 124 | \n",
" 155 | \n",
" 69.0 | \n",
" 924059 | \n",
"
\n",
" \n",
" 13 | \n",
" 14 | \n",
" Female | \n",
" 140 | \n",
" 120 | \n",
" 147 | \n",
" 155 | \n",
" 70.5 | \n",
" 856472 | \n",
"
\n",
" \n",
" 14 | \n",
" 15 | \n",
" Female | \n",
" 96 | \n",
" 100 | \n",
" 90 | \n",
" 146 | \n",
" 66.0 | \n",
" 878897 | \n",
"
\n",
" \n",
" 15 | \n",
" 16 | \n",
" Female | \n",
" 83 | \n",
" 71 | \n",
" 96 | \n",
" 135 | \n",
" 68.0 | \n",
" 865363 | \n",
"
\n",
" \n",
" 16 | \n",
" 17 | \n",
" Female | \n",
" 132 | \n",
" 132 | \n",
" 120 | \n",
" 127 | \n",
" 68.5 | \n",
" 852244 | \n",
"
\n",
" \n",
" 17 | \n",
" 18 | \n",
" Male | \n",
" 100 | \n",
" 96 | \n",
" 102 | \n",
" 178 | \n",
" 73.5 | \n",
" 945088 | \n",
"
\n",
" \n",
" 18 | \n",
" 19 | \n",
" Female | \n",
" 101 | \n",
" 112 | \n",
" 84 | \n",
" 136 | \n",
" 66.3 | \n",
" 808020 | \n",
"
\n",
" \n",
" 19 | \n",
" 20 | \n",
" Male | \n",
" 80 | \n",
" 77 | \n",
" 86 | \n",
" 180 | \n",
" 70.0 | \n",
" 889083 | \n",
"
\n",
" \n",
" 20 | \n",
" 21 | \n",
" Male | \n",
" 83 | \n",
" 83 | \n",
" 86 | \n",
" NaN | \n",
" NaN | \n",
" 892420 | \n",
"
\n",
" \n",
" 21 | \n",
" 22 | \n",
" Male | \n",
" 97 | \n",
" 107 | \n",
" 84 | \n",
" 186 | \n",
" 76.5 | \n",
" 905940 | \n",
"
\n",
" \n",
" 22 | \n",
" 23 | \n",
" Female | \n",
" 135 | \n",
" 129 | \n",
" 134 | \n",
" 122 | \n",
" 62.0 | \n",
" 790619 | \n",
"
\n",
" \n",
" 23 | \n",
" 24 | \n",
" Male | \n",
" 139 | \n",
" 145 | \n",
" 128 | \n",
" 132 | \n",
" 68.0 | \n",
" 955003 | \n",
"
\n",
" \n",
" 24 | \n",
" 25 | \n",
" Female | \n",
" 91 | \n",
" 86 | \n",
" 102 | \n",
" 114 | \n",
" 63.0 | \n",
" 831772 | \n",
"
\n",
" \n",
" 25 | \n",
" 26 | \n",
" Male | \n",
" 141 | \n",
" 145 | \n",
" 131 | \n",
" 171 | \n",
" 72.0 | \n",
" 935494 | \n",
"
\n",
" \n",
" 26 | \n",
" 27 | \n",
" Female | \n",
" 85 | \n",
" 90 | \n",
" 84 | \n",
" 140 | \n",
" 68.0 | \n",
" 798612 | \n",
"
\n",
" \n",
" 27 | \n",
" 28 | \n",
" Male | \n",
" 103 | \n",
" 96 | \n",
" 110 | \n",
" 187 | \n",
" 77.0 | \n",
" 1062462 | \n",
"
\n",
" \n",
" 28 | \n",
" 29 | \n",
" Female | \n",
" 77 | \n",
" 83 | \n",
" 72 | \n",
" 106 | \n",
" 63.0 | \n",
" 793549 | \n",
"
\n",
" \n",
" 29 | \n",
" 30 | \n",
" Female | \n",
" 130 | \n",
" 126 | \n",
" 124 | \n",
" 159 | \n",
" 66.5 | \n",
" 866662 | \n",
"
\n",
" \n",
" 30 | \n",
" 31 | \n",
" Female | \n",
" 133 | \n",
" 126 | \n",
" 132 | \n",
" 127 | \n",
" 62.5 | \n",
" 857782 | \n",
"
\n",
" \n",
" 31 | \n",
" 32 | \n",
" Male | \n",
" 144 | \n",
" 145 | \n",
" 137 | \n",
" 191 | \n",
" 67.0 | \n",
" 949589 | \n",
"
\n",
" \n",
" 32 | \n",
" 33 | \n",
" Male | \n",
" 103 | \n",
" 96 | \n",
" 110 | \n",
" 192 | \n",
" 75.5 | \n",
" 997925 | \n",
"
\n",
" \n",
" 33 | \n",
" 34 | \n",
" Male | \n",
" 90 | \n",
" 96 | \n",
" 86 | \n",
" 181 | \n",
" 69.0 | \n",
" 879987 | \n",
"
\n",
" \n",
" 34 | \n",
" 35 | \n",
" Female | \n",
" 83 | \n",
" 90 | \n",
" 81 | \n",
" 143 | \n",
" 66.5 | \n",
" 834344 | \n",
"
\n",
" \n",
" 35 | \n",
" 36 | \n",
" Female | \n",
" 133 | \n",
" 129 | \n",
" 128 | \n",
" 153 | \n",
" 66.5 | \n",
" 948066 | \n",
"
\n",
" \n",
" 36 | \n",
" 37 | \n",
" Male | \n",
" 140 | \n",
" 150 | \n",
" 124 | \n",
" 144 | \n",
" 70.5 | \n",
" 949395 | \n",
"
\n",
" \n",
" 37 | \n",
" 38 | \n",
" Female | \n",
" 88 | \n",
" 86 | \n",
" 94 | \n",
" 139 | \n",
" 64.5 | \n",
" 893983 | \n",
"
\n",
" \n",
" 38 | \n",
" 39 | \n",
" Male | \n",
" 81 | \n",
" 90 | \n",
" 74 | \n",
" 148 | \n",
" 74.0 | \n",
" 930016 | \n",
"
\n",
" \n",
" 39 | \n",
" 40 | \n",
" Male | \n",
" 89 | \n",
" 91 | \n",
" 89 | \n",
" 179 | \n",
" 75.5 | \n",
" 935863 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Unnamed: 0 Gender FSIQ VIQ PIQ Weight Height MRI_Count\n",
"0 1 Female 133 132 124 118 64.5 816932\n",
"1 2 Male 140 150 124 NaN 72.5 1001121\n",
"2 3 Male 139 123 150 143 73.3 1038437\n",
"3 4 Male 133 129 128 172 68.8 965353\n",
"4 5 Female 137 132 134 147 65.0 951545\n",
"5 6 Female 99 90 110 146 69.0 928799\n",
"6 7 Female 138 136 131 138 64.5 991305\n",
"7 8 Female 92 90 98 175 66.0 854258\n",
"8 9 Male 89 93 84 134 66.3 904858\n",
"9 10 Male 133 114 147 172 68.8 955466\n",
"10 11 Female 132 129 124 118 64.5 833868\n",
"11 12 Male 141 150 128 151 70.0 1079549\n",
"12 13 Male 135 129 124 155 69.0 924059\n",
"13 14 Female 140 120 147 155 70.5 856472\n",
"14 15 Female 96 100 90 146 66.0 878897\n",
"15 16 Female 83 71 96 135 68.0 865363\n",
"16 17 Female 132 132 120 127 68.5 852244\n",
"17 18 Male 100 96 102 178 73.5 945088\n",
"18 19 Female 101 112 84 136 66.3 808020\n",
"19 20 Male 80 77 86 180 70.0 889083\n",
"20 21 Male 83 83 86 NaN NaN 892420\n",
"21 22 Male 97 107 84 186 76.5 905940\n",
"22 23 Female 135 129 134 122 62.0 790619\n",
"23 24 Male 139 145 128 132 68.0 955003\n",
"24 25 Female 91 86 102 114 63.0 831772\n",
"25 26 Male 141 145 131 171 72.0 935494\n",
"26 27 Female 85 90 84 140 68.0 798612\n",
"27 28 Male 103 96 110 187 77.0 1062462\n",
"28 29 Female 77 83 72 106 63.0 793549\n",
"29 30 Female 130 126 124 159 66.5 866662\n",
"30 31 Female 133 126 132 127 62.5 857782\n",
"31 32 Male 144 145 137 191 67.0 949589\n",
"32 33 Male 103 96 110 192 75.5 997925\n",
"33 34 Male 90 96 86 181 69.0 879987\n",
"34 35 Female 83 90 81 143 66.5 834344\n",
"35 36 Female 133 129 128 153 66.5 948066\n",
"36 37 Male 140 150 124 144 70.5 949395\n",
"37 38 Female 88 86 94 139 64.5 893983\n",
"38 39 Male 81 90 74 148 74.0 930016\n",
"39 40 Male 89 91 89 179 75.5 935863"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import pandas\n",
"data = pandas.read_csv('examples/brain_size.csv', sep=';', na_values=\".\")\n",
"data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
">**分割符**\n",
"它是CSV文件,但是分割符是”;”\n",
"\n",
">**缺失值**\n",
"CSV中的第二个个体的weight是缺失的。如果我们没有指定缺失值 (NA = not available) 标记符, 我们将无法进行统计分析。\n",
"\n",
"**从数组中创建**: [pandas.DataFrame](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html#pandas.DataFrame) 也可以视为1D序列, 例如数组或列表的字典,如果我们有3个`numpy`数组:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import numpy as np\n",
"t = np.linspace(-6, 6, 20)\n",
"sin_t = np.sin(t)\n",
"cos_t = np.cos(t)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"我们可以将他们暴露为[pandas.DataFrame](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html#pandas.DataFrame):"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"
\n",
" \n",
" \n",
" | \n",
" cos | \n",
" sin | \n",
" t | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" 0.960170 | \n",
" 0.279415 | \n",
" -6.000000 | \n",
"
\n",
" \n",
" 1 | \n",
" 0.609977 | \n",
" 0.792419 | \n",
" -5.368421 | \n",
"
\n",
" \n",
" 2 | \n",
" 0.024451 | \n",
" 0.999701 | \n",
" -4.736842 | \n",
"
\n",
" \n",
" 3 | \n",
" -0.570509 | \n",
" 0.821291 | \n",
" -4.105263 | \n",
"
\n",
" \n",
" 4 | \n",
" -0.945363 | \n",
" 0.326021 | \n",
" -3.473684 | \n",
"
\n",
" \n",
" 5 | \n",
" -0.955488 | \n",
" -0.295030 | \n",
" -2.842105 | \n",
"
\n",
" \n",
" 6 | \n",
" -0.596979 | \n",
" -0.802257 | \n",
" -2.210526 | \n",
"
\n",
" \n",
" 7 | \n",
" -0.008151 | \n",
" -0.999967 | \n",
" -1.578947 | \n",
"
\n",
" \n",
" 8 | \n",
" 0.583822 | \n",
" -0.811882 | \n",
" -0.947368 | \n",
"
\n",
" \n",
" 9 | \n",
" 0.950551 | \n",
" -0.310567 | \n",
" -0.315789 | \n",
"
\n",
" \n",
" 10 | \n",
" 0.950551 | \n",
" 0.310567 | \n",
" 0.315789 | \n",
"
\n",
" \n",
" 11 | \n",
" 0.583822 | \n",
" 0.811882 | \n",
" 0.947368 | \n",
"
\n",
" \n",
" 12 | \n",
" -0.008151 | \n",
" 0.999967 | \n",
" 1.578947 | \n",
"
\n",
" \n",
" 13 | \n",
" -0.596979 | \n",
" 0.802257 | \n",
" 2.210526 | \n",
"
\n",
" \n",
" 14 | \n",
" -0.955488 | \n",
" 0.295030 | \n",
" 2.842105 | \n",
"
\n",
" \n",
" 15 | \n",
" -0.945363 | \n",
" -0.326021 | \n",
" 3.473684 | \n",
"
\n",
" \n",
" 16 | \n",
" -0.570509 | \n",
" -0.821291 | \n",
" 4.105263 | \n",
"
\n",
" \n",
" 17 | \n",
" 0.024451 | \n",
" -0.999701 | \n",
" 4.736842 | \n",
"
\n",
" \n",
" 18 | \n",
" 0.609977 | \n",
" -0.792419 | \n",
" 5.368421 | \n",
"
\n",
" \n",
" 19 | \n",
" 0.960170 | \n",
" -0.279415 | \n",
" 6.000000 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" cos sin t\n",
"0 0.960170 0.279415 -6.000000\n",
"1 0.609977 0.792419 -5.368421\n",
"2 0.024451 0.999701 -4.736842\n",
"3 -0.570509 0.821291 -4.105263\n",
"4 -0.945363 0.326021 -3.473684\n",
"5 -0.955488 -0.295030 -2.842105\n",
"6 -0.596979 -0.802257 -2.210526\n",
"7 -0.008151 -0.999967 -1.578947\n",
"8 0.583822 -0.811882 -0.947368\n",
"9 0.950551 -0.310567 -0.315789\n",
"10 0.950551 0.310567 0.315789\n",
"11 0.583822 0.811882 0.947368\n",
"12 -0.008151 0.999967 1.578947\n",
"13 -0.596979 0.802257 2.210526\n",
"14 -0.955488 0.295030 2.842105\n",
"15 -0.945363 -0.326021 3.473684\n",
"16 -0.570509 -0.821291 4.105263\n",
"17 0.024451 -0.999701 4.736842\n",
"18 0.609977 -0.792419 5.368421\n",
"19 0.960170 -0.279415 6.000000"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pandas.DataFrame({'t': t, 'sin': sin_t, 'cos': cos_t})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**其他输入**: [pandas](http://pandas.pydata.org/) 可以从SQL、excel文件或者其他格式输入数。见[pandas文档](http://pandas.pydata.org/)。\n",
"\n",
"#### 3.1.1.2.2 操作数据\n",
"\n",
"`data`是[pandas.DataFrame](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html#pandas.DataFrame), 与R的dataframe类似:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(40, 8)"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.shape # 40行8列"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"Index([u'Unnamed: 0', u'Gender', u'FSIQ', u'VIQ', u'PIQ', u'Weight', u'Height',\n",
" u'MRI_Count'],\n",
" dtype='object')"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.columns # 有列"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0 Female\n",
"1 Male\n",
"2 Male\n",
"3 Male\n",
"4 Female\n",
"5 Female\n",
"6 Female\n",
"7 Female\n",
"8 Male\n",
"9 Male\n",
"10 Female\n",
"11 Male\n",
"12 Male\n",
"13 Female\n",
"14 Female\n",
"15 Female\n",
"16 Female\n",
"17 Male\n",
"18 Female\n",
"19 Male\n",
"20 Male\n",
"21 Male\n",
"22 Female\n",
"23 Male\n",
"24 Female\n",
"25 Male\n",
"26 Female\n",
"27 Male\n",
"28 Female\n",
"29 Female\n",
"30 Female\n",
"31 Male\n",
"32 Male\n",
"33 Male\n",
"34 Female\n",
"35 Female\n",
"36 Male\n",
"37 Female\n",
"38 Male\n",
"39 Male\n",
"Name: Gender, dtype: object\n"
]
}
],
"source": [
"print(data['Gender']) # 列可以用名字访问"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"109.45"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# 简单选择器\n",
"data[data['Gender'] == 'Female']['VIQ'].mean()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"> **注意**: 对于一个大dataframe的快速预览,用它的`describe`方法: [pandas.DataFrame.describe()](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.describe.html#pandas.DataFrame.describe)。\n",
"\n",
"**groupby**: 根据类别变量的值拆分dataframe:"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"('Female', 109.45)\n",
"('Male', 115.25)\n"
]
}
],
"source": [
"groupby_gender = data.groupby('Gender')\n",
"for gender, value in groupby_gender['VIQ']:\n",
" print((gender, value.mean()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**groupby_gender**是一个强力的对象,暴露了结果dataframes组的许多操作:"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"
\n",
" \n",
" \n",
" | \n",
" Unnamed: 0 | \n",
" FSIQ | \n",
" VIQ | \n",
" PIQ | \n",
" Weight | \n",
" Height | \n",
" MRI_Count | \n",
"
\n",
" \n",
" Gender | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
" | \n",
"
\n",
" \n",
" \n",
" \n",
" Female | \n",
" 19.65 | \n",
" 111.9 | \n",
" 109.45 | \n",
" 110.45 | \n",
" 137.200000 | \n",
" 65.765000 | \n",
" 862654.6 | \n",
"
\n",
" \n",
" Male | \n",
" 21.35 | \n",
" 115.0 | \n",
" 115.25 | \n",
" 111.60 | \n",
" 166.444444 | \n",
" 71.431579 | \n",
" 954855.4 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Unnamed: 0 FSIQ VIQ PIQ Weight Height MRI_Count\n",
"Gender \n",
"Female 19.65 111.9 109.45 110.45 137.200000 65.765000 862654.6\n",
"Male 21.35 115.0 115.25 111.60 166.444444 71.431579 954855.4"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"groupby_gender.mean()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"在`groupby_gender`上使用tab-完成来查找更多。其他的常见分组函数是median, count (对于检查不同子集的缺失值数量很有用) 或sum。Groupby评估是懒惰模式,因为在应用聚合函数之前不会进行什么工作。\n",
"\n",
">**练习**\n",
"- 完整人口VIO的平均值是多少?\n",
"- 这项研究中包含了多少男性 / 女性?\n",
"- **提示** 使用‘tab完成’来寻找可以调用的方法, 替换在上面例子中的‘mean’。\n",
"- 对于男性和女性来说,以log为单位显示的MRI count平均值是多少?"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![](http://www.scipy-lectures.org/_images/plot_pandas_1.png)\n",
"\n",
">**注意**: 上面的绘图中使用了`groupby_gender.boxplot` (见[这个例子](http://www.scipy-lectures.org/packages/statistics/auto_examples/plot_pandas.html#example-plot-pandas-py))。\n",
"\n",
"#### 3.1.1.2.3 绘制数据\n",
"\n",
"Pandas提供一些绘图工具 (`pandas.tools.plotting`, 后面使用的是matplotlib) 来显示在dataframes数据的统计值:\n",
"\n",
"**散点图矩阵**:"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([[,\n",
" ,\n",
" ],\n",
" [,\n",
" ,\n",
" ],\n",
" [,\n",
" ,\n",
" ]], dtype=object)"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/matplotlib/collections.py:590: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
" if self._edgecolors == str('face'):\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEjCAYAAAD31uwUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGIZhGC3z/wHvRjfgFyU/MAAAAABJRU5ErkJggg==\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from pandas.tools import plotting\n",
"plotting.scatter_matrix(data[['Weight', 'Height', 'MRI_Count']]) "
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([[,\n",
" ,\n",
" ],\n",
" [,\n",
" ,\n",
" ],\n",
" [,\n",
" ,\n",
" ]], dtype=object)"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAERCAYAAAB8eMxzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJztnXuUHVWV/z87dIfQiCBqIPkhg4jjYAKZGAdQOkACRgV5\nDIsREZzJMI4IM5oZETDiK+gaBIMafj8IDjqABBUXChFQ0yTpTBIGBBuMDYgoEBwggBAIj4akb7J/\nf1Td7tuV+6h7b1Xd1/ezVq+uqlunzrm179l1ap+99zF3RwghROcwrtENEEIIkS1S/EII0WFI8Qsh\nRIchxS+EEB2GFL8QQnQYUvxCCNFhpKb4zWySmd1jZq+a2bjw2CYz6zezlWa2W3jsVDO73cxuNrNd\n0mqPEEKIAEvLj9/MdgR2Am4EjnT3bWa2xt1nFpzTDawAjgBOAvZ294WpNEgIIQSQ4ojf3Te7+wuR\nw/ub2WozuzDcfzsw6O7bgOXAe9JqjxBCiICujOvbz91fMLMrzOxY4FngxfCzF4HdogXMTKHFQghR\nA+5uxY5nqvgL3gBuAqYDS4HXh8deD0TfEAD48pe/PLJ9xBFHcMQRR6TXyDbAzIC4z0sjSXPfqlWr\nWLVq1cj+ggULEr1+u9DXt5Krr34SgLlzJzNnzuwGt2h7zJL9bYjaqeX3EuiBEp+lLVgz6weOAnYE\nNrv7VjP7GrCOwP6/AphFCRu/mbl+fNXRSMVfrC2dIL/h4WH6+9cAMGvWTLq7u8t+DpQ9vxnoFNk1\nmkq/ncJzcrkcAF1dXRV/N6H8sh3xm1kX8EtgWvj/fGCxmb0MPAJ80d3dzK4E1gAbgY+m1R4h0qS/\nf83IiAzWbDciK/Z5M47yRfZU+u0AdHd3M2fO7DEj/1LnxiE1xe/uOYKRfiEzipy3BFiSVjuEEEKM\nJXVTT73I1FM9MvVkT7WmnmY07UTpFNk1mmp+G9WcW87Uk6Yf/yTgVmB/YOfQZRMzOxH4trvvHe6f\nCpxFaOpx95ci15HirxIpfpEEkl1rU07xp5myYSMwG7gzcvwk4E9hw7qBM4CZwLXhthBCiBTJNIDL\nzI4GbmN0OJpKAJeZVfUnhBCdRNZJ2v6esRO5u1EhgKt2POafEEJ0FpkpfjObDdzh7sMFh18gRgCX\nEEKI5MgqcteAqcBxZvYBYIqZXQBcAEwNs3ceBdxRrPBXvvKVkW1F7jY30chd0Vy0oneRqEy1ck3T\nqycfwPUu4B5gvrvfHX622t0PC7dPA84kQa+eZvJqaQTN9P3lGdJcVBP6L9m1DsXk2pDI3RIBXPnP\nDivYVgCXEEJkSFsGcDXTiLcRNNP316ixNtLK7VNDAFBN9YjaKJRPb+8hrF0beMPXEtjVqFw9YwK4\ngN0JsnEOE6RjPsXdhysFcAnRiUTztwCJ5GjJ53wRzUmh3AcGFjM4uEf4SXmZVyvXNCd38wFcN+b3\n3f1QADP7EvAhM7uF0QCuk8JtrcBVAsUctC/FRviiMyiUfT77ZtqkaePfDGzOK6t8yoaQ1wHPAfsR\nBnCZ2XLgyrTa0z7EefXWA6LViI7wA+UffRBs/2Co1ptDXj3NR6HsTzttInPnTgagt/dE1q69k1wu\nRy6Xo69v5YjM6pVjpguxmNlBwGUE/vrzgYNJLYBLiNal2Kt7sVf5OCl96zlfZEtXV9cYmZRKxVyv\nHLNegesu4G/M7DPA6cDtKIBLiO1G+ENDQyxatBiAefPOpKenJ5F6crkcGzY8FW5PHPOZzE2N4eCD\nZ/CjH90Qbl9Y9Jxt23I8/fRDDAw8zaxZM8vKMQ6ZBXCZWXdB1O5LwHjgIRTA1VYogKs2oiP8hQsv\n5YorhsK9xcyff3bRcsVNQpXYWPRosQllkT6XX/5dVqyYNLIdlfWsWTMZGLiUvr7dWbfuzSMP51Jy\njEOmK3CZ2cUERupNwKnunouzAleh4hfNTfTBvGDBgsY1pgOo2pujq4tJk945si2an+7ubmbMmM7g\n4JMjx+qVo/z4W8iPP/73ap7vL1/w2kjL1FNuUjD62fjx4yW7DIgj61riOhrixy+EiE8l+3pS3jjl\n3hCGh4cZGLgXCIKHRPIUC9DK5XJMm3bAyMi9r28lMFbOebkNDQ2xcOGlQH0DgiwDuP4CuIZgKPo4\n8LHQjVMBXKLjidrXBwbuHWPjnzFjeureOIsWLR5Tp0ieYgFawSTtRiZNemfFoK2ojErN/VQiyxW4\nngeOcffDgUeBo7UClxBCZE+WAVyFrprDQA4FcAkBbO+dE5haRu2+wSt/uq6W8+adOabOz3/+s6nU\n08kUynk0QCtwx+zq6ho5NnruWKIyqpXUJ3fNrB84smCx9cnAj4HDgEOAY919fugFtMzdj4yU1+Ru\niCZ3O5dGRNxKdumStkybZnLXzHYErgY+Ho7ytQKXEDFQxG370UiZZrkCF8B/Ape5+4Ph/h9QAFdb\noQAuIZqfWKYeM9sZ6I7Y6SuVKVyBawD4KnBLuA3wbXdfqhW44iNTT+ciU0/70UhTT1nFb2ZnAecB\nu4SHhoBL3P1bZranuz+VaEuLt0GKP0SKX2SJZNfa1GTjN7MvAO8FZrn7I+GxfYFFZvYW4HBgRgrt\nFUIIkSIlR/xm9hAwzd1fjRzfCfgzQa6dpak3UCP+ETTiF1ki2bU25Ub85QK4tkWVPkB47IlKSt/M\nJpnZPWb2qpmNM7NuM7vDzF4K3xzy551qZreb2c1mtku5awohhKifcor/STM7KnrQzI4Enohx7TGR\nu2FK5uOBGwi9fBS5K4QQ2VPOnfNTwFIzW0vgiWMENv1e4LhKF45G7obHnomsG/t2FLkrEqCSh0Ra\n2S5FdpSSsWRbPSUVv7vfb2YHEOTIn0JgNF4NfLKYCahGdkVLL4oEqBQMk1RyK9E4SslYsq2esgFc\noYL/Xgr15meMNhEjclcBXK2DAriEaH7KefW8TGnXEHf315f4LHqdfuAod98a7l8FfM3dHw5t/MuB\nWcBJwN7uvjBSXl49IfLqKY1MPcnTbF49MvVUR80BXHVWGo3cPR84m2CO4FHgIne/WZG78ZHiF1ki\n2bU2NSn+0F//k8DbgEHge+6eS62VJWgmxR+ZmC5LGh1Gil9kiWTX2tTqx38NgRfPfcDRwCUptK0F\n8Rh/QgjRvJSb3N3f3Q8AMLPvAndXc+Ho0ouhy+Y5BK6gjwFz3T2npRdFtdSS3KoRSc5EaWqVh+SY\nDOUU/4hZJ1TQ1V47H8B1I4CZTQSOcPeZZnYucIKZLWU0gOukcHthiesJAdSWx1z57JuLWuUhOSZD\nOcV/oJkVjr53Ktiv6NUTCeAy4N3AqvDj5cCpwP0ogEsIITKlXADXDgnXVSxYa7cix4QoS3R92rTK\niPSoVR6SYzJktQKXEwRr7RXu54O1FMDVZmQRwNXd3b3dK34l22+xMqJx5OVRrc1eckyGrBZbPwp4\nI/Bf7v6h0Mb/CHATsIIWCeCqxp1S7pzh1TNyCezrWzli+507d7KUQwJkITvJLT0asth6QQDXtPD/\n+cBqM1tD4NXzzXDS+EqCd7eNBHmBMqeGietEr9kKvtLV3qPC79TXtxKQF0a7Ik+b1iP1EX+9ZDHi\nT/7cxr9xJF1/rW9RZsYpp1wLJDeiK6ZopHySJ+6Iv9KovZxsJLf0aMiIX4i0KObSJ9tv81LOBVNy\nawyZKv7Q/LMEmAjc7e7nFQvqyrJNIn3mzp0MyAujXZGnTeuRqanHzP4O2NfdLzKzSwlW4zrP3Y/J\nT/i6+w2RMjL15M9sUVNP0r8xmQeyISnZSV6NoSHZOUs05FzgPnf/uZmdDkwAetx9oZm9i2AB97Mj\nZaT482dK8SeKFFJ5Gik7yaZ+msnG/3vgcODnBOkcHgCeCz9TAJeoi2qVhcL/s6Ua+Ug26ZK14r8Z\nODJMz7CeIGBLAVxtxSpgrMyyQsqiuZF8modMFb+7bwM+DWBm3wFuAS4HvkEQ5HVHsXJf+cpXmDPn\nOFatWsF//Mc3smquqIkjgFHFv2DBgsY1pQKalGxeJJt0ydrGPxm4DtgGXOPu3w/t/sdSwqsnb+M/\n+OD3c9ddZxE8H8rxNMHaMbLxJ1l/Wjb+OK//cU0E8hdPllKyi95LoCr55HJBF+/q6pIsUqRpbPzu\n/iRBeobCYxcDF8e7wgRg5wrnaL3NViLO63+cc4opo8KI4XrMDHqgjCV6L4Ex+/n7DcXvybp1g6xb\n92bGjRtHJVmUur+deN+TRAFcoi2opIySvHahopLdentK3ZP88Q0bngOcSZP2rPta0eMiHlkHcO0I\n/JhgIncT8GFgHgrg6lji2HKTsPfKZpwcxe/l6H5+JF6KPfb4S6ZNe44ZMyZLFg0i6xH/Bwgidr9m\nZp8HTiGyKhdBUJfoEOKE7Mc5p7f3EAYGFofbJ4av/mNNAbWOCss9NDrxgRK9l8PDw2M+L3VPxh4/\nNZZ5Jt61OuO+J0nWk7uHAn/r7p81s28Q+O6/GieAK5jc/Qzw/gq1PAVMQpO7rTG5mxRK75s8SSVp\nE42h3OTuuIzbcgfwLjO7D5gB/BHIL+eoAC4hhMiArE09HwNudfdLzOxsoJuYAVyPP/5H4FpgR/K+\n4qIZWQVkH8ClV//GoXvfemRt6vkU8Iq7/5eZzQX2Bg4qXJWrVJI2mXrSM/XEp3lNPSJ5JLvWpmn8\n+AlSMl9vZh8DtgAnA58oXJUr4/YIIP7DRAjRDmQdwPU8MCdyuIoALiGKo4CebNH9bm0UwCXaAgX0\nZIvud2sjxS+aHo0uG4fufXsixS+anjijS3mWpEOpe6/73dqkpvjNbBJwK7A/sLO7bzOzTcA9BLOJ\nJ7r7C2Z2KnAWsBH4qLu/VPKiQpRAi3Zni+53a5PmiH8jwSpbNxYc+627j2TnNLNu4AxgJnBSuL2w\n9CWXAn+oUO2LtbVWNC0aXTYO3fv2JDXF7+6bgc2B7/kI+5vZauB2d58PvB0YDN8GlgNXFrtW5Box\nqaZMGufGv2Z13y/5+tP+TrXJTzQD48ePb3QTRApkbePfLzTvXGFmxwLPMjpEL5myQUEkpbnwwku4\n4oohAD75yR7mzz+7qXKnKAioeqoJ1Evz3kp2zUe+v7/22ib22OMVpk49tGQfLzfgytqPP5+S4SZg\nOoHtRmvuthGrVq1i1apVjW6GEKIMqadsMLN+gvUSdwQ2u/tWM/sasI7A/r+CYFWuk4C93X1hpLxr\n1FGaoaEhFi0K0hHPm3cmPT09TeWCp1Fj9WjEL0qR7+9bt27lwAOnMmHChJJ9vCEpG8ysC/glMC38\nfz6w2MxeBh4BvujubmZXEswebQQ+mlZ72pWensC8U45KD4JmelAIIbansI9+9rOfBhiz4E21fTjN\nyd0c26+MPqPIeUsIcviIhKi0DGHUHqgoTCGamzhLi1bTh7POxy+EEKLBZBrAFR4/Efi2u+8d7iuA\nK2EqrYka73whRLMQr0/H78OpTe6GC6vvRDCBe2SB4v8BwSRubxjAtYJgZRVN7rYhmiCsHk3uiiRo\nyNKL7r65wH0z35CjgdsY/VWPBHABy4H3pNUe0Z6YWew/IURA1jb+v2fsRO5uxAjgEqI8HuNPCJEn\nswAuM5sN3OHuwwWjrxdQAFdboQAuIZqfLAO4/gU4nmDJxYOB/wdcgAK4UqeRfvpp24mbxR6eJM3y\nnWTjbw5q7b/NEMC1DJjv7peGn6129y+F2wrgShn56QvRuqTRf7MO4Mp/dljBtgK4hBAiQ1I39dSL\nTD31I1MPyNRTWzta5Z61M2mYetL04x8TwAXsTpCNc5ggHfMp4URv2QAuKf6AUsKv5UeR5YNAir96\nmuU7SfGnS74f5nI5ALq6uuru24U0xMbP9itwbXT3Q8MGfQn4kJndQlUrcHUupex8tdj/ZPMXovHk\n++GGDU8BG5k06Z3U27fjklkAVz5yN+R1wHPAfiiASwghMiXThVjM7CDgMgJ//fkEbp0K4IpBqXw6\nBx88gx/96IZw+0KgeI7+Qnp7D2FgYHG4fWLaTRdNQLWRyzLxpE++H+6332s8+uif6e7+Hw444Itc\neOElAJx11sdHzk06h1bWK3DdBfyNmX0GOB24HQVwxaK7u7voq97ll3+XFSsmjWzPn382ixYtHlmO\nERZvl69/7do7GRzcY2Q7yVdIBXA1M3GVudJbZEG+H9533+08/fSuTJiwK088cSYPPpjPXv/dimtt\n1EpWit/MrNvdh8P9l4DxwEPAVDMbR+D6eUexwoWKX4xl69atvPrqo+H2X5U9Nz9ZNDBwL9u2vZFx\n45IXf/TBvGDBgsTrEKJVKZywzU/qum9j8+ZXANi2bVvJskmS6QpcZnYxwbBjE3Cqu+cUwFUfBx44\nlUmTNo1sQ2DegcUF2wH5yaJt297MtGl/ZsaM6UrDLESGFE7YnnbaRObOncwdd+zDT37yGPAKn/jE\nXP74x2AgV9h3kybrAK4jipynAK46mDBhAlOm7D+yDZWXYxw3bhwzZkyXN48QDaSrq2ukD/7+928B\nYNddd0vNvFOIArhanGp8fdP23y92ffnxV08a3yn+NUevKz/+5CnWR+L0y1r6brMEcP0FcA3Br+9x\n4GPuvk0BXO1DX9/KkdfYuXMnM2fObCn+GpDiF1GK9a1KNGQhFkYDuO4M958HjnH3w4FHgaPDFbjy\nAVzXhttCCCFSJE0b/2Zgc95/OLIa1zCQoyCAy8yWA1em1R6RPNHXT63dK0Q65PtWLpcjl8vR17ey\nLnNtpn78AGY2GXgf8FXgEBTA1bIUCylv5gnjaoKYZOIQzUQ+jqfQ5FNPGoesI3d3BK4GPh6O8rUC\nV5vR3AFcCmASAjJcgcvdt5rZNcBP3X1p+Fk3QY4ercDVgsTxNGimyd1qJzcbhSZ3RSmq8e6p2avH\nzI4GzgOmAN3Ab4FvuvuNZrZL1AMnUjYfwPUuYIDAtHNLuA3wbXdfamanAWcir562RIq/eqT4RRLU\npPjN7J8JvGzOZVRZzwC+DlwHnODus5Jv7nbtkOJvUjTiTwcp/s4irfiaWvPxfwbodffnCo6tNLMP\nAk8A/16h0qgf/w7AamAqMM3dHwnPK+vHL5oX5fUXon4a0Y/K+vFHlH7hsfXuvrjCtcf48YcJ2o4H\nbiCcPcvKj394eJi+vpX09a1keHi4cgEhRFOjPl0f5Ub8L5rZX7v7bwoPmtk0giRrZYn68YfHnom4\n1L2dDPz4NTJNB/nti0bRTn26Ef2onOI/G1hqZlcR2PiNwMY/Fzgtofp3RX78LUupNQKEEPFpRD8q\nqfjdfa2ZHQz8C4GyB3gAONjdn6qz3vyM0SZi+PHXi0amrUW1q0W1Au34nRqJ+nR9lA3gChX8FxOo\nJ/qrz+//gSoXYlEAV3OTXABXOwZbteN3SpdSHi9626yPcu6cg2XKubsfWPbC2/vxn09gPuolSNJ2\nkbvfnIUffy2Z7UQy1OISWK3rYSu4c6b1ndrdnVN9t3Zqdef8X+DC8L9T5TCkxEIsJxc5TwuxCCFE\nhpRT/H3AxcBk4Hrgh+5+byatShjZA4VoTdR306Firh4z2wf4CMFovQf4AcFD4KEK5cYEcIUum+cA\nxwGPAXPDNXe1EEsbI1NPeEWZekTG1LUQi7uvd/evu/t0ggfA3wK/i1HvmAAuM5sIHOHuMwly/pyg\nhViEECJ7Kip+M+sys+PM7AcEk7UPAidWKufumwsWXzHg3cCqcH858B4KFmIpOCaEECJFStr4zWwO\nwQj/GOAu4IfAJ9z95RrrKhastVuRY5mT9iLkQojqUJ9Ml3KTu58jUPafdfeNddbjBMFae4X7+WCt\nTAK4KtFO4d9CtAPqk+lSLnI3yTttwK8JJnG/wWiw1kMogKutaO4VuIQQkOIKXCUCuI4AjmWsV0+m\nC7EUe4VM4rWymmsMDQ2xaFGQ3HTevDPp6empur5WQV494RXl1TNCnL5S6pxixzupP1VDrQFcdVEi\ngOsugtiAwvMyDeAq9gqZRPh3Na+mixYt5oorhsK9xcyff3ZddQvRSsTpK6X6ZLGy6k/VU9GrRwgh\nRHuR+mLrYyoLzD9LgInA3e5+XrGgrkiZuk09ha+Hvb2HsHbtnUCy3gIy9RRHpp7wijL1jPSRXC7o\n4l1dXVX3QZl64lPzYuspNOTvgH3d/SIzu5RgNa7z3P0YMzsXeMTdb4iUUZK2FkaKP7yiFL/6YcbU\nFbmbMG8F8lk/f0Ow/m5/uN+QAC4t4SZENuRyOTZseIANGx4YGfWLxpDa5G4Jfg8cDvycIJ3DA0B+\nXd9YAVy1eAT09h7CwEDwKtjbOzbouFp/4aw9gIRoRsqZT4Eyv+/dS16nUl9Qv0mOrBX/zcCR4fq6\n6wkCtqoK4IqjqKPnAAwO7gHA2rV31vWKmURgSacFpzz77LN8+tPnoBeq9qHwNzwwsHikf+X7W7Hf\nd1dXF5Mm7TmyHb1Opb7Qaf0mTTJV/GFOnk8DmNl3gFuAyxkb1LUdhQFcO+/8eoK54WRQ2tdkKRbA\n9corr/DTn/6MzZsXxrjC0lTalRZaUjE+6mvNQ9aTu5OB64BtwDXu/v1wUndMUFekzJjJ3VpMPVDu\n1bM01QR76ZW1OGbG+vXrmTLlMF555bEYJS4AvkyrTO7GnzRtr8ndakw90f3CflSNh08n9ZskaBqv\nnlpoZD7+arwQ5LFQHCn+2upvdsVfDcX6hvpL+jSTV48QQogGk6mN38x2BH5MMJG7CfgwMI8yAVyQ\n7CteNdeqxiYp+6VoR+rJq5OnWN9Qf2ksWXv1fIAgYvdrZvZ54BTCVblCW/8JBEFdY0hyNr+aa1WT\nwyeJfD9CNBu1eNFFzynWN9RfGkvWpp5nGfXVfwOwNw0O4BJCiE4j6xH/HcBXzew+4BngSkb9+EsG\ncCX5WqhXTCHiE6e/qE+1Hlkr/o8Bt7r7JWZ2NtBNjACuJF8L9YopRHzi9Bf1qdYja8X/euD5cPs5\nYB/gIKoI4Kp3BS75AqdLu6zA1SqBWY1op/pQ65N1ANcbgOsJRvpbgJOBT1BFAFe9yH84W1rVj781\n6k/vO5Xz41cfag0asgJXMdz9eWBO5PDFRFblEkIIkR5Zm3oajiaihKgP9aHWJ+sArvcDnwt330Gw\nyPpfUiGAK0k0ESVEfagPtT6Z+vG7+zJ3n+Xus4A/AfcQBnABvyUI4BJCCJEiDcnVY2b7Ak8TrMC1\nKjzctAFcWqVLiNKof7QejbLxnwj8lCBg68XwWKwVuBqBFoAQojTqH61Ho7Jzfgj4GUGitqpW4BKt\nihN48Fb601qsQqRN5vn4zWxP4PvuPsfMJgL/5e4fCpO0PeLuN0TOb+4FA4QQoklppnz8xwE3Abj7\nM8BqM1sDHJg/HsXdq/778pe/nEmZrMulUdeyZSs45ZRrOeWUa1m2bEWidQFN8z1VV2lZFytXqe/V\n+l2SKN/IupMqH2q3mH9UXX85UlP8ZjbJzO4xs1fNrLCeZxl16QR4ImzHrsBOabVHCCFEQJqTuxuB\n2cCNkeMnEbhyYmbdwBnAzPD4GUCcFblFGYqtOfzww4/S17cy9kIZoj3p7T2EgYHF4faJYz7L/24e\nfvhReee0OakpfnffDGwuTCJlZkcDtwGnh4feDgy6+zYzW06QpjkRaknkVmvytyzLRcsMDQ2xaFHQ\nkefNO5Oenp7tvCwAnnxyr/BYvIUy4rSvVLKuaLlml0UWdeXv1c47v57h4eGqE5sl9b3Wrr2TwcE9\nRrYL5f6LX9zGF75wMy+//Dy/+MVtqbQpqfKNrDup8gsWLGhY/alP7ppZP3BkqNx/RJCaeaUHq269\nFzjW3eebWRewzN2PjJT3tNvYylx44SVcccUQAJ/8ZA/z55+9XRItIJWkWnGSdSW1YHer0yyJzcq1\n4/TTz+L663cB4OSTX+KqqxZLdikSDIqrS5xX7fW90UnazGw2cIe7Dxe8BbxADHfOJNMydwLR1/lg\ndJm+KWd4eJhvf/v/8tvfrmOfffZmhx12SK2uVqdRqY3LmXr23XcfXve6Z0a2RfuSleI3gijd48zs\nA8AUM7uAIAfv1HDyN1Y+fjGWefPOBBYXbG//Op+Wso/ODfT3r+HeeycC72PmzGA0Wc/rbDtR7F4l\nEfRU7QOknKnnM5/5V3bYYfS39MUvnldTm0Tzk5riD003vwSmAcuA+e5+afjZanf/Urh9JUGP2Ah8\nNK32tCs9PYF5pxxRJZNXPFDfaFPJuuJT770qpeCTjJqN81sSjSPuojtxTEJpTu7mCEbxxT47rGB7\nCbAkrXZ0IsVGl4WkFWIv76D4VHuvkpKZZNTKxF2cpzIdl4+/3Sg2EoyOLqOd/bbb+tmw4QEAcrmJ\nibVFbwDxqXSvot5apahWkUtGAqT4W544I8HinX331NsmamfRosUj3lqwmM9+9tMUU/BS5KIWpPhb\nnFwux4YNT4Xbwei9mG8/jL4drFs3yB57vIlx47ro6qr9J6BFt0ep915Ey2/ZsoVNm34PwJYtU5Jt\nrOh40pzcnQTcCuwP7EwwxFwKDBOkbTgldO08FTiLcHLX3V9Kq02tTCnFksvleO6534Tb7wS2Hy3m\nJ+zybwfbtr2ZadP+zIwZ0+uy8yod7yj13oto+a1btzI8PB6ArVu3cttt/XzjG6uAQOZHHx1dulqI\n+KSZpC2fsuHO/L67H+ruRxCstvWhSMqGa8NtUYS8Yrj66ifHTNbef//v2LTpHWza9A7uv/93sa41\nbtw4ZsyYzpw5szt6lN7MPP74k8CbgTfz+ONPsm7dII88shOPPLIT69YNNrp5osXJLGWDu28r+Ph1\nwHPAfqSUsqFTmDJlf3bd9eZw+1iguG8/JO/RIQ+RUeq9F9HAqtdee41f/zqQ6wknHEtXVxf77vtb\nAKZNOyCZRouOJevF1g8CLiOI0J0PHEwLrMDVDJRSLF1dXbzxjW8b2YbS/thJTwRqYnGUeu9FNLDq\ngx98HxMmTABG5Z2Xb6c/ZEX9ZKr43f0u4G/M7DMEidpuRykbYlFKsXR1dTFp0p4j241m1apVrFq1\nqtHNaHmKyVsPWZEUWSVpOwoY5+7D4bF/BsYD3wFWALMI0jLv7e4LI+WVpK0Mze5ZoyRt8WhGOUp2\n6VJtkray+hFEAAAR40lEQVS4AVx5mZVL0paa4i9I2fAuYAA4H7iYoPWbgFPd/RUzOw04kxJePVL8\nrY2UR+si2aVLWyr+pJDib22kPFoXyS5dGqn4G7HmrhBCiAYixS+EEB1GZoutm9lbzWy1mf23mV2X\nX4DdzE41s9vN7GYz2yWt9gghhAjIMnL3eeAYdz8ceBQ4WpG7QgiRPakpfnff7O4vFOy/UOCxMwzk\nKIjcBZYD70mrPUIIIQIyt/Gb2WTgfUAf8AYUuSuEEJmSqeI3sx2Bq4GPh6P8WIutCyGESI6yMf5m\ndjRwHjAF6CbIqvlNd7/RzHapIoVy3pf0P4HL3P3BcP8PVLnYeqembGgVlLJBiOanZABXmFbhDOBc\ngshbgBnA14HrgBPcfVbJC28fuftV4JaCa33b3Zcqcre9URBQ65KX3cMPP8z69etjlzv00ENHEsyJ\n0jRl5K6Z/Q7odffnIsffCDwB/Lu7L47Z6pqR4m9tpPhbl7zs5s8/n29964dMmPDWimVefnkt69c/\nzF577ZVBC1ubRir+sqaeqNLPHzOz9VkofSFE43GHzZv/ic2bz694bk+PFH4rUG5y90Uz++voQTOb\nRpBkTQghRAtSbsR/NrDUzK4isMsbgY1/LnBa+k0TQgiRBiVH/O6+lmCFrB0IlP0/hOcf7O5rSpXL\nUyRlQ7eZ3WFmL5nZvgXnKWWDEEJkSCUb/1PAF2u8dj5lw43htYbN7HjgIkL3zkjKhpPC7YVFryaE\nECIRSip+MxssU87d/cByF44uth4ee6ZwH3g7WmxdCCEypdyI/9gM6t8VpWwQQohMKan43X194b6Z\nvQk4DHjM3QeKFopP3iF1E1psva1Q5K54y1veEvvcuDEeEUtBItfsZMqZem4FznP3+8xsEnAvcDfw\nNjO70t2/VUU9Uanl96tO2SCam+iDecGCBY1rjGgg8YKNGn/NzqScH/8+7n5fuP2PQJ+7H0vg6XN6\npQubWVdot58G/NLMDjKz64E5wDVmdqy7DxPY9dcAHwO+U8d3EUIIEYNyNv7hgu2jCCde3f0lM9tW\n6cLungvLFXJykfOWAEsqN1UIIUQSlFP8j5vZpwjy8kwnSLiGmfVUKCeEEKKJKWfq+SdgKkHg1snu\n/nx4/GDgqrQbJoQQIh3Kjdx3dPft1sB1936gP70mtT9DQ0MsWhTkuJs370x6enoa3KL2R/dciFHK\njfiX5jfM7CfVXjiasiE8do6ZrTGzJWG+/o5M2bBo0WKuuGKIK64YVUYiXXTPmxMzi/UnkiXu0ov7\nVj5lO/IpG+4EMLOJwBHuPpNgJa8TIikbrg23E2d4eJi+vpX09a1keHi4cgEhYqLfVr14zD+RJKlN\n0kZSNhjwbmBV+PFy4FTgfjJI2dDfv4arr34y3FvDnDmzE69jeHiY/v4gd92sWTPp7u4uee68eWcC\niwu2RdqcddbH+cMf5ofbFyZ23Sx+W0IkTTnFf6CZ5ZdB3KlgG4JcPa8vVqgMxdIz7FbkWEtSjQLo\n7u5mxozpI9uifio9eH/1qwFee+29I9tS0KKTKZeyYYcE63GC9Az55Xny6RkySdkwa9ZMghix/HZj\naedRYqNSNjTqnjbbb0uIOGTlj2/Ar4GzgG8wmp7hIdokZYMUQECzpmyoRz7l3ia6u7vb6sEtOoPU\nFH/otfNLwpQNwPnAajNbAzwGfNPdc2aWT9mwEfhoGm2pNBqsxj5fimoUgB4SyVPpnhaTT1y5t/Mb\nmuhM0pzcLZay4S7g4sh5DU/ZkHXH1igxeWq5p1LoolPpiNQLGmGLetDvR7QbHaH4K40G1bE7k7hy\n1xuaaDcyVfyh3X8JMBG4293PM7NzgOMI7P5zQxNRplTTsZOYDxDJU4tcpNBFp5L1iP9vgXvd/SIz\nu9TMDiOM5jWzc4ETgBsyblNVlLIL64GQHcXutez1QsQna8X/ViC/iPtvCLJ/5hO+5aN5U1X8aSlo\nKZ7sqOVe68EsxChZK/7fA4cDPyfI4/MA8Fz4WWqRu4WdPpfLsWTJM+Ena5g1a2ZVCkHzAc3JrFkz\nyeX6WbdukFxuIkNDQ6xde+fIZ3owCzFK1or/ZuDIMC/PeoJI3dQjdws7/QEHPA3sUfSzqEIoNkos\nZRfWAyEgi8jd3t5DGBhYHG6fCAT2+q6uLgYH92Bw8BnWrVvM4GBezmtSbY8QrUamit/dtwGfBjCz\n7wC3AJczNpp3O5KM3J027QBmzAi+duFovxjV5t/RKDKbyN21a+8cUepr194Z677rwSzEKFl79UwG\nrgO2Ade4+5/MbEw0bxr1Rjt9oTlHCqF9KJRlb++JY0w9ejALMYq5N3euazPzRrVRE4L1Y2YkLT/J\nJRvysvvc587noot6CLKulKenZy+Ghp4gXg59i3leNecm/3tLiyBlfXrfP5Rf0VVsOiKAq1ZFoVFi\nc1JP3h0hRIcofnl0tD+SsRDxaUvFHx39FZLL5ejrWznyWZyRoUaT9VHt/Y4S5/7ncjk2bHgg3J5Y\nR2uFaH+yntzdEfgxgevmJuDDwDwSTtkQHf0VTvrlcrmqR4YaTdZHvfcu/v3fvZbmCdFxxF1sPSk+\nQJCjZxZBiuZTiCzAXukCtSxunbcJz5kzm66utnzJEUCwpMPGRjdCdDBmFvuvkWStBZ9lNDr3DcAO\nVJmyIc7or5yLZrHgn0rI5bM+5s6dDBS/d3HMOPHvv0b8ohmoxlOnMWSt+O8Avmpm9wHPAFcyGrmb\nWMqGct44tQT/yLunPsrduzgP8jj3v6uri0mT9hzZFkKUJuse8jHgVne/xMzOBrqpMmVDb29v2RGk\naCyNWmxdb2VCVIG7Z/YHfAo4PdyeC3wJuCXcPxc4qUgZ37Jliy9btsKXLVvhW7Zs8Tj09/cXPV7u\nWqXK1FpXGuVarS6g7HVKyaOwTDXy74R7mlVdgXpwP++8zzt8zcEjf/3bHevp+T8OFDm32F/c84qd\nu33d+fPSvDeVysf/Tv2pf/9wu6guznpydwnwETPrJ5jYvZTRBdgPBG4qVihvDrj66ifL5tYppNSo\ns3CiN2pPrnWkmmW5Vqyr3HVKyaOwTDXy75R72hx11Xb9ZKiv7nrfSut/q21s+ayTtD0PzIkcvpjI\nAuxCCCHSoyVmwWS/7Wwk/8ZiBt3dP2H8+D+OOb5ly72MH//YmGObN8udthVoiSRtjW6DEEK0Il4i\nSVvTK34hhBDJkvXkrhBCiAYjxS+EEB2GFL8QQnQYTeXVY2ZdBIna3kOQvuEFgjQPN3mJrJ21lKmj\nrl2AM4qU+Y67v1SmrszKNUEbNwEvhf93JiUZ1iH3pr6nzVhXrfc6wfI13ZN2KF9v3SUpFdnViD+C\nAK9zgXcB+4X/zwWWJFmmjrpuJkglvTvBQ3P3cP/mCnVlVq4J2vjj8N6uTlOGdci9qe9pM9ZV671O\nsHxN96Qdytdbd8nr1lM46T9gTTXHay1TR123A+Mix8YBt1eoK7NyjW4jgcP9duWSlmEdcm/qe9qM\nddV6rxMsX9M9aYfy9dZd6q+pTD3Az8zsVoJ45BcJErcdTvDUq1Smn8DEEKdMreUuB1aZ2WBB+6YC\niyvUlWW5RrdxHPBnYJmZnUF6MqxV7s1+T5uxrlrvdVLla70n7VC+3rqL0lR+/GZ2PHAPwRfL24t/\n7e7PlCkzniDvz97AHwkyfu4DXOruRbN9FpQ9rKCuFwgSxv2pQplugtfVfPse8nh2ym7g7cCuYbk/\nuHvFlWQK6ntD2MaK5YqUqbaN+ftRy3ebAdwC7E/KMqxFfgXtrUqG7Sq/It9ruzL19LGCa9Qkq0g7\nq+53kfJVy69I/bHlWKF87PbX2i/LXrPJFP+TwJ+Ap4GfAj/zIL9PuTI3Eazm9QYCpXMr8Bxwiru/\nv0y5i4CdgN8As4HXgK0Er1DfL1GmrkmqItc71t1LjnrMbBxwNJAD+tx9W3j8eHdfWqJMN/Ahgntw\nO3A6MB642t1fqbJ9F7j7lyqcs1u+85vZscB1wFPA70hRhrXILyyXmAxbXX5FZDeVQLHf4AWKodY+\nVlC+JlkVlE+03xVct6z8Cs6rWo6R8onJNLxexX5Z8RpNpvj73X2Wme0LnEhws7YQCPjycmXC7fvc\nfWr0eIlyK919dsH+cnc/ysxWuPuRJcosIVgicjmjr11HAQe6+2ll6tqXscvteLh/jbv3lil3HfAo\nMAy8D/i4uz9Y7ruZ2Q1hG18HTAd+QfB6fYy7l1za0sz+RPDQLfxBTAHuc/fDypRb6e6zzexCAsXw\nHoJX+HcAvyIlGdYiv/C8qmXYrvIrIrubgF7g/7j7PxacV1Mfi9ZTsB9LVgXn19TvCsrXJL+C8lXL\nMVK+JpmGZWvql5VoNhs/AO7+CLAQWGhmexIsxl6KF83sCwQjgWcsWODleQIhleMZMzsPGCSwNz4Q\nHi8X2/AXRX5o91iQVroc6yi+pORbK5Tby91PBTCzK4GrzeyyCmXe5O4XhGXudvdvhtsfrVDu34CT\ngNsIvC2GzewX7v7BCuXyHepQdz8s7AxfMLP/dvc0ZViL/KA2Gbar/MbILtz+pZn9d+S8WvtYnlpl\nlafWfpenVvnlqUWOhdQqU6i9X5bH65gZTvoPeH8NZXYEjgcOJfgh/QPBgi9vqFCuC/g7Arey4wln\nzoHJZcqcQ/Caew6Bb+05BPbscyvUdQcwscjx6yuUWw3sEvmu1wAvlilze8H2QQXbq2Lez6OBHwHz\ngBUxzt9E4MmTX0/5/QRrKQ+kKcNa5FerDNtVflHZhce2k12tfaxeWdUjsyTkV48ck5RptXKN89dU\npp5WwMwmAu9mdJLo18A+7n5XmTLdXmQiyMwOqlDurcDzHplAM7PD3H11iTJvAjZ6aIfM1w9ML1dX\nkevMJrD53uHud8ctF5btAaZWU1+WVCvDTpJfs8quln5XULYm+RWcV7UcI+clItOwXM39cgz1Pjk6\n6Y9gtDOOYFS0Q8H2bTHLjYtcI/FyKdS1PI3v1koybFf5tYrsau13SX3PRpavtV9W+mtKG38T8wpw\nZ5Hj05qoXCu0sZE0+z1thbqypt52tnL5dGTU6Kd5K/0RxBjsVuR4pVFxZuVaoY2tJsN2lV+ryK7e\ndrZy+bRkJBt/FZjZJAJb3ebI8S4vnxgss3Kt0MZG0uz3tBXqypp629nK5dOSkRS/EEJ0GHH9aIUQ\nQrQJUvxCCNFhSPELIUSHIcWfMma21czuNbNBM/uxme0UHn+54JwpZrbSzB40s4fC8HjRYEKZzIkc\n+zcz+7kFaXLzx3rN7Fdm9rvw75+zb63IU9Dn8n97m9keZnaLmf3GzO63IE00ZrZPJ8pSij99htx9\nursfQJCs7JPhcQcIHwRLgf9w978i8M99r5md1ZDWikJ+CHwkcuxk4ML8TpiH6DrgDHffnyDJ2Rlm\ndnRmrRRR8n0u//cn4AJgmbv/tbtPAc6LFuokWUrxZ8ta4G2RYx8F1rr7cgB3fxX4V4r8MEXm/AQ4\nxoK0wJjZPsBk4H8LzvkX4Cp3/w2Auz9HkJNG8msu9gSeyO+4+31FzukYWUrxZ0SoPD5IkKGwkCnA\nQOEBD7KT7mJmO2fUPFEEd99IkIc+P+L7CHA9Y1PkvpOI/ML9Kak3UJRipwIzz0/CY5cB3wvNd58P\n/eOjdIwslbIhfXYys3vD7dXA9yKf53ODF6M7tVaJuOTNPT8jMPOczvbyKiY/ya5xvOru0wsPuHuf\nBXn5P0AwALvXzKYWKdsRstSIP31eLbA1zisSbfcAwapGI4Q/0CGPuaydSJWfAUea2XSgx93vjXy+\nnfzC/d9m0TgRH3d/3t1/6O5/D9wNRBcy6RhZSvE3nh8AvWZ2JIxM9i4CLmloqwQA7v4ywSLhVxHI\nKsplwFwzmwZgZm8Evg58M7NGioqY2aww5TRmtgvBXNtjkdM6RpZS/OlTKieGw8hk7nHA+Wb2e+DP\nBAs5fyuj9onK/BA4IPyfJy+/p4DTgP80swcJJhCvc/cbM2+lyFOsz80A7jazdcD/AFe6+0Dh+Z0k\nS+XqaTLM7HjgSwTrcT7V6PaI6jCzM4EPA8e5+0uNbo+onXaWpRS/EEJ0GDL1CCFEhyHFL4QQHYYU\nvxBCdBhS/EII0WFI8QshRIchxS+EEB2GFL8QQnQY/x9496oyV9p1VgAAAABJRU5ErkJggg==\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plotting.scatter_matrix(data[['PIQ', 'VIQ', 'FSIQ']]) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
">**两个总体**\n",
"\n",
"> IQ指标是双峰的, 似乎有两个子总体。\n",
"\n",
"> **练习**\n",
"\n",
"> 只绘制男性的散点图矩阵,然后是只有女性的。你是否认为2个子总体与性别相关?\n",
"\n",
"## 3.1.2 假设检验: 比较两个组\n",
"\n",
"对于简单的[统计检验](https://en.wikipedia.org/wiki/Statistical_hypothesis_testing),我们将可以使用[scipy](http://docs.scipy.org/doc/)的子摸块[scipy.stats](http://docs.scipy.org/doc/scipy/reference/stats.html#module-scipy.stats):"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"from scipy import stats"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"> **也看一下**: Scipy是一个很大的库。关于整个库的快速预览,可以看一下[scipy](http://nbviewer.ipython.org/github/cloga/scipy-lecture-notes_cn/blob/master/1.5.%20Scipy%EF%BC%9A%E9%AB%98%E7%BA%A7%E7%A7%91%E5%AD%A6%E8%AE%A1%E7%AE%97.ipynb) 章节。\n",
"\n",
"### 3.1.2.1 Student’s t检验: 最简单的统计检验\n",
"\n",
"#### 3.1.2.1.1 单样本 t-检验: 检验总体平均数的值\n",
"\n",
"[scipy.stats.ttest_1samp()](http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.ttest_1samp.html#scipy.stats.ttest_1samp)检验数据总体的平均数是否可能等于给定值 (严格来说是否观察值来自于给定总体平均数的正态分布)。它返回一个[T统计值](https://en.wikipedia.org/wiki/Student%27s_t-test)以及[p-值](https://en.wikipedia.org/wiki/P-value) (见函数的帮助):"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(30.088099970849328, 1.3289196468728067e-28)"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_1samp(data['VIQ'], 0) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"根据$10^-28$的p-值,我们可以声称IQ(VIQ的测量值)总体平均数不是0。\n",
"\n",
"![](http://www.scipy-lectures.org/_images/two_sided.png)\n",
"\n",
"#### 3.1.2.1.2 双样本 t-检验: 检验不同总体的差异\n",
"\n",
"我们已经看到男性和女性总体VIQ平均数是不同的。要检验这个差异是否是显著的,我们可以用[scipy.stats.ttest_ind()](http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.ttest_ind.html#scipy.stats.ttest_ind)进行双样本检验:"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(-0.77261617232750113, 0.44452876778583217)"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"female_viq = data[data['Gender'] == 'Female']['VIQ']\n",
"male_viq = data[data['Gender'] == 'Male']['VIQ']\n",
"stats.ttest_ind(female_viq, male_viq)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1.2.2 配对实验: 同一个体的重复测量\n",
"\n",
"PIQ、VIQ和FSIQ给出了IQ的3种测量值。让我检验一下FISQ和PIQ是否有显著差异。我们可以使用双样本检验:"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(0.46563759638096403, 0.64277250094148408)"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_ind(data['FSIQ'], data['PIQ'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"使用这种方法的问题是忘记了两个观察之间有联系: FSIQ 和 PIQ 是在相同的个体上进行的测量。因此被试之间的差异是混淆的,并且可以使用\"配对实验\"或\"[重复测量实验](https://en.wikipedia.org/wiki/Repeated_measures_design)\"来消除。"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(1.7842019405859857, 0.082172638183642358)"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_rel(data['FSIQ'], data['PIQ'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![](http://www.scipy-lectures.org/_images/plot_paired_boxplots_1.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"这等价于单样本的差异检验:"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(1.7842019405859857, 0.082172638183642358)"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_1samp(data['FSIQ'] - data['PIQ'], 0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![](http://www.scipy-lectures.org/_images/plot_paired_boxplots_2.png)\n",
"\n",
"T-tests假定高斯误差。我们可以使用[威尔科克森符号秩检验](https://en.wikipedia.org/wiki/Wilcoxon_signed-rank_test), 放松了这个假设:"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(274.5, 0.10659492713506856)"
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.wilcoxon(data['FSIQ'], data['PIQ']) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"> **注意: **非配对实验对应的非参数检验是[曼惠特尼U检验](https://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U), [scipy.stats.mannwhitneyu()](http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.mannwhitneyu.html#scipy.stats.mannwhitneyu)。\n",
"\n",
"> **练习**\n",
"- 检验男性和女性重量的差异。\n",
"- 使用非参数检验来检验男性和女性VIQ的差异。\n",
"\n",
">**结论**: 我们发现数据并不支持男性和女性VIQ不同的假设。"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3.1.3 线性模型、多因素和因素分析\n",
"\n",
"### 3.1.3.1 用“公式” 来在Python中指定统计模型\n",
"\n",
"#### 3.1.3.1.1 简单线性回归\n",
"\n",
"给定两组观察值,x和y, 我们想要检验假设y是x的线性函数,换句话说:\n",
"\n",
"$y = x * coef + intercept + e$\n",
"\n",
"其中$e$是观察噪音。我们将使用[statmodels module](http://statsmodels.sourceforge.net/):\n",
"- 拟合一个线性模型。我们将使用简单的策略,[普通最小二乘](https://en.wikipedia.org/wiki/Ordinary_least_squares) (OLS)。\n",
"- 检验系数是否是非0。\n",
"\n",
"![](http://www.scipy-lectures.org/_images/plot_regression_1.png)\n",
"\n",
"首先,我们生成模型的虚拟数据:"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import numpy as np\n",
"x = np.linspace(-5, 5, 20)\n",
"np.random.seed(1)\n",
"# normal distributed noise\n",
"y = -5 + 3*x + 4 * np.random.normal(size=x.shape)\n",
"# Create a data frame containing all the relevant variables\n",
"data = pandas.DataFrame({'x': x, 'y': y})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"> **Python中的统计公式**\n",
"\n",
"> [见statsmodels文档](http://statsmodels.sourceforge.net/stable/example_formulas.html)\n",
"\n",
"\n",
"然后我们指定一个OLS模型并且拟合它:"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"from statsmodels.formula.api import ols\n",
"model = ols(\"y ~ x\", data).fit()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"我们可以检查fit产生的各种统计量:"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: y R-squared: 0.804\n",
"Model: OLS Adj. R-squared: 0.794\n",
"Method: Least Squares F-statistic: 74.03\n",
"Date: Wed, 18 Nov 2015 Prob (F-statistic): 8.56e-08\n",
"Time: 17:10:03 Log-Likelihood: -57.988\n",
"No. Observations: 20 AIC: 120.0\n",
"Df Residuals: 18 BIC: 122.0\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"==============================================================================\n",
" coef std err t P>|t| [95.0% Conf. Int.]\n",
"------------------------------------------------------------------------------\n",
"Intercept -5.5335 1.036 -5.342 0.000 -7.710 -3.357\n",
"x 2.9369 0.341 8.604 0.000 2.220 3.654\n",
"==============================================================================\n",
"Omnibus: 0.100 Durbin-Watson: 2.956\n",
"Prob(Omnibus): 0.951 Jarque-Bera (JB): 0.322\n",
"Skew: -0.058 Prob(JB): 0.851\n",
"Kurtosis: 2.390 Cond. No. 3.03\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
]
}
],
"source": [
"print(model.summary())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"> **术语**:\n",
"\n",
"> Statsmodels使用统计术语: statsmodel的y变量被称为‘endogenous’而x变量被称为exogenous。更详细的讨论见[这里](http://statsmodels.sourceforge.net/devel/endog_exog.html)。\n",
"\n",
"> 为了简化,y (endogenous) 是你要尝试预测的值,而 x (exogenous) 代表用来进行这个预测的特征。\n",
"\n",
"> **练习**\n",
"\n",
"> 从以上模型中取回估计参数。**提示**: 使用tab-完成来找到相关的属性。\n",
"\n",
"#### 3.1.3.1.2 类别变量: 比较组或多个类别\n",
"\n",
"让我们回到大脑大小的数据:"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"data = pandas.read_csv('examples/brain_size.csv', sep=';', na_values=\".\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"我们可以写一个比较,用线性模型比较男女IQ:"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: VIQ R-squared: 0.015\n",
"Model: OLS Adj. R-squared: -0.010\n",
"Method: Least Squares F-statistic: 0.5969\n",
"Date: Wed, 18 Nov 2015 Prob (F-statistic): 0.445\n",
"Time: 17:34:10 Log-Likelihood: -182.42\n",
"No. Observations: 40 AIC: 368.8\n",
"Df Residuals: 38 BIC: 372.2\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"==================================================================================\n",
" coef std err t P>|t| [95.0% Conf. Int.]\n",
"----------------------------------------------------------------------------------\n",
"Intercept 109.4500 5.308 20.619 0.000 98.704 120.196\n",
"Gender[T.Male] 5.8000 7.507 0.773 0.445 -9.397 20.997\n",
"==============================================================================\n",
"Omnibus: 26.188 Durbin-Watson: 1.709\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 3.703\n",
"Skew: 0.010 Prob(JB): 0.157\n",
"Kurtosis: 1.510 Cond. No. 2.62\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
]
}
],
"source": [
"model = ols(\"VIQ ~ Gender + 1\", data).fit()\n",
"print(model.summary()) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"---\n",
"\n",
"**特定模型的提示**\n",
"\n",
"**强制类别**: ‘Gender’ 被自动识别为类别变量,因此,它的每一个不同值都被处理为不同的实体。\n",
"使用:"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"model = ols('VIQ ~ C(Gender)', data).fit()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"可以将一个整数列强制作为类别处理。\n",
"\n",
"**截距**: 我们可以在公式中用-1删除截距,或者用+1强制使用截距。\n",
"\n",
"默认,statsmodel将带有K和可能值的类别变量处理为K-1'虚拟变量' (最后一个水平被吸收到截距项中)。在绝大多数情况下,这都是很好的默认选择 - 但是,为类别变量指定不同的编码方式也是可以的 (http://statsmodels.sourceforge.net/devel/contrasts.html)。\n",
"\n",
"**FSIQ和PIQ差异的t-检验**\n",
"\n",
"要比较不同类型的IQ,我们需要创建一个\"长形式\"的表格,用一个类别变量来标识IQ类型:"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" iq type\n",
"0 133 fsiq\n",
"1 140 fsiq\n",
"2 139 fsiq\n",
"3 133 fsiq\n",
"4 137 fsiq\n",
"5 99 fsiq\n",
"6 138 fsiq\n",
"7 92 fsiq\n",
"8 89 fsiq\n",
"9 133 fsiq\n",
"10 132 fsiq\n",
"11 141 fsiq\n",
"12 135 fsiq\n",
"13 140 fsiq\n",
"14 96 fsiq\n",
"15 83 fsiq\n",
"16 132 fsiq\n",
"17 100 fsiq\n",
"18 101 fsiq\n",
"19 80 fsiq\n",
"20 83 fsiq\n",
"21 97 fsiq\n",
"22 135 fsiq\n",
"23 139 fsiq\n",
"24 91 fsiq\n",
"25 141 fsiq\n",
"26 85 fsiq\n",
"27 103 fsiq\n",
"28 77 fsiq\n",
"29 130 fsiq\n",
".. ... ...\n",
"10 124 piq\n",
"11 128 piq\n",
"12 124 piq\n",
"13 147 piq\n",
"14 90 piq\n",
"15 96 piq\n",
"16 120 piq\n",
"17 102 piq\n",
"18 84 piq\n",
"19 86 piq\n",
"20 86 piq\n",
"21 84 piq\n",
"22 134 piq\n",
"23 128 piq\n",
"24 102 piq\n",
"25 131 piq\n",
"26 84 piq\n",
"27 110 piq\n",
"28 72 piq\n",
"29 124 piq\n",
"30 132 piq\n",
"31 137 piq\n",
"32 110 piq\n",
"33 86 piq\n",
"34 81 piq\n",
"35 128 piq\n",
"36 124 piq\n",
"37 94 piq\n",
"38 74 piq\n",
"39 89 piq\n",
"\n",
"[80 rows x 2 columns]\n"
]
}
],
"source": [
"data_fisq = pandas.DataFrame({'iq': data['FSIQ'], 'type': 'fsiq'})\n",
"data_piq = pandas.DataFrame({'iq': data['PIQ'], 'type': 'piq'})\n",
"data_long = pandas.concat((data_fisq, data_piq))\n",
"print(data_long) "
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: iq R-squared: 0.003\n",
"Model: OLS Adj. R-squared: -0.010\n",
"Method: Least Squares F-statistic: 0.2168\n",
"Date: Wed, 18 Nov 2015 Prob (F-statistic): 0.643\n",
"Time: 18:16:40 Log-Likelihood: -364.35\n",
"No. Observations: 80 AIC: 732.7\n",
"Df Residuals: 78 BIC: 737.5\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"===============================================================================\n",
" coef std err t P>|t| [95.0% Conf. Int.]\n",
"-------------------------------------------------------------------------------\n",
"Intercept 113.4500 3.683 30.807 0.000 106.119 120.781\n",
"type[T.piq] -2.4250 5.208 -0.466 0.643 -12.793 7.943\n",
"==============================================================================\n",
"Omnibus: 164.598 Durbin-Watson: 1.531\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 8.062\n",
"Skew: -0.110 Prob(JB): 0.0178\n",
"Kurtosis: 1.461 Cond. No. 2.62\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
]
}
],
"source": [
"model = ols(\"iq ~ type\", data_long).fit()\n",
"print(model.summary()) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"我们可以看到我们获得了与前面t-检验相同的值,以及相同的对应iq type的p-值:"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(0.46563759638096403, 0.64277250094148408)"
]
},
"execution_count": 32,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_ind(data['FSIQ'], data['PIQ'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1.3.2 多元回归: 包含多因素\n",
"\n",
"考虑用2个变量x和y来解释变量z的线性模型:\n",
"\n",
"$z = x \\, c_1 + y \\, c_2 + i + e$\n",
"\n",
"这个模型可以被视为在3D世界中用一个平面去拟合 (x, y, z) 的点云。\n",
"\n",
"![](http://www.scipy-lectures.org/_images/plot_regression_3d_1.png)\n",
"\n",
"**实例: 鸢尾花数据 ([examples/iris.csv](examples/iris.csv))**\n",
"\n",
"萼片和花瓣的大小似乎是相关的: 越大的花越大! 但是,在不同的种之间是否有额外的系统效应?\n",
"\n",
"![](http://www.scipy-lectures.org/_images/plot_iris_analysis_1.png)"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: sepal_width R-squared: 0.478\n",
"Model: OLS Adj. R-squared: 0.468\n",
"Method: Least Squares F-statistic: 44.63\n",
"Date: Thu, 19 Nov 2015 Prob (F-statistic): 1.58e-20\n",
"Time: 09:56:04 Log-Likelihood: -38.185\n",
"No. Observations: 150 AIC: 84.37\n",
"Df Residuals: 146 BIC: 96.41\n",
"Df Model: 3 \n",
"Covariance Type: nonrobust \n",
"======================================================================================\n",
" coef std err t P>|t| [95.0% Conf. Int.]\n",
"--------------------------------------------------------------------------------------\n",
"Intercept 2.9813 0.099 29.989 0.000 2.785 3.178\n",
"name[T.versicolor] -1.4821 0.181 -8.190 0.000 -1.840 -1.124\n",
"name[T.virginica] -1.6635 0.256 -6.502 0.000 -2.169 -1.158\n",
"petal_length 0.2983 0.061 4.920 0.000 0.178 0.418\n",
"==============================================================================\n",
"Omnibus: 2.868 Durbin-Watson: 1.753\n",
"Prob(Omnibus): 0.238 Jarque-Bera (JB): 2.885\n",
"Skew: -0.082 Prob(JB): 0.236\n",
"Kurtosis: 3.659 Cond. No. 54.0\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
]
}
],
"source": [
"data = pandas.read_csv('examples/iris.csv')\n",
"model = ols('sepal_width ~ name + petal_length', data).fit()\n",
"print(model.summary()) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1.3.3 事后假设检验: 方差分析 (ANOVA)\n",
"\n",
"在上面的鸢尾花例子中,在排除了萼片的影响之后,我们想要检验versicolor和virginica的花瓣长度是否有差异。这可以被公式化为检验在上面的线性模型中versicolor和virginica系数的差异 (方差分析, ANOVA)。我们写了\"差异\"向量的参数来估计: 我们想要用[F-检验](https://en.wikipedia.org/wiki/F-test)检验 \"`name[T.versicolor] - name[T.virginica]`\":"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n"
]
}
],
"source": [
"print(model.f_test([0, 1, -1, 0]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"是否差异显著?\n",
"\n",
"> **练习**\n",
"> 回到大脑大小 + IQ 数据, 排除了大脑大小、高度和重量的影响后,检验男女的VIQ差异。\n",
"\n",
"## 3.1.4 更多可视化: 用seaborn来进行统计学探索\n",
"\n",
"[Seaborn](http://stanford.edu/~mwaskom/software/seaborn/) 集成了简单的统计学拟合与pandas dataframes绘图。\n",
"\n",
"让我们考虑一个500个个体的工资及其它个人信息的数据 ([Berndt, ER. The Practice of Econometrics. 1991. NY: Addison-Wesley](http://lib.stat.cmu.edu/datasets/CPS_85_Wages))。\n",
"\n",
"加载并绘制工资数据的完整代码可以在[对应的例子](http://www.scipy-lectures.org/packages/statistics/auto_examples/plot_wage_data.html#example-plot-wage-data-py)中找到。"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" EDUCATION SOUTH SEX EXPERIENCE UNION WAGE AGE RACE OCCUPATION \\\n",
"0 8 0 1 21 0 5.10 35 2 6 \n",
"1 9 0 1 42 0 4.95 57 3 6 \n",
"2 12 0 0 1 0 6.67 19 3 6 \n",
"3 12 0 0 4 0 4.00 22 3 6 \n",
"4 12 0 0 17 0 7.50 35 3 6 \n",
"5 13 0 0 9 1 13.07 28 3 6 \n",
"6 10 1 0 27 0 4.45 43 3 6 \n",
"7 12 0 0 9 0 19.47 27 3 6 \n",
"8 16 0 0 11 0 13.28 33 3 6 \n",
"9 12 0 0 9 0 8.75 27 3 6 \n",
"10 12 0 0 17 1 11.35 35 3 6 \n",
"11 12 0 0 19 1 11.50 37 3 6 \n",
"12 8 1 0 27 0 6.50 41 3 6 \n",
"13 9 1 0 30 1 6.25 45 3 6 \n",
"14 9 1 0 29 0 19.98 44 3 6 \n",
"15 12 0 0 37 0 7.30 55 3 6 \n",
"16 7 1 0 44 0 8.00 57 3 6 \n",
"17 12 0 0 26 1 22.20 44 3 6 \n",
"18 11 0 0 16 0 3.65 33 3 6 \n",
"19 12 0 0 33 0 20.55 51 3 6 \n",
"20 12 0 1 16 1 5.71 34 3 6 \n",
"21 7 0 0 42 1 7.00 55 1 6 \n",
"22 12 0 0 9 0 3.75 27 3 6 \n",
"23 11 1 0 14 0 4.50 31 1 6 \n",
"24 12 0 0 23 0 9.56 41 3 6 \n",
"25 6 1 0 45 0 5.75 57 3 6 \n",
"26 12 0 0 8 0 9.36 26 3 6 \n",
"27 10 0 0 30 0 6.50 46 3 6 \n",
"28 12 0 1 8 0 3.35 26 3 6 \n",
"29 12 0 0 8 0 4.75 26 3 6 \n",
".. ... ... ... ... ... ... ... ... ... \n",
"504 17 0 1 10 0 11.25 33 3 5 \n",
"505 16 0 1 10 1 6.67 32 3 5 \n",
"506 16 0 1 17 0 8.00 39 2 5 \n",
"507 18 0 0 7 0 18.16 31 3 5 \n",
"508 16 0 1 14 0 12.00 36 3 5 \n",
"509 16 0 1 22 1 8.89 44 3 5 \n",
"510 17 0 1 14 0 9.50 37 3 5 \n",
"511 16 0 0 11 0 13.65 33 3 5 \n",
"512 18 0 0 23 1 12.00 47 3 5 \n",
"513 12 0 0 39 1 15.00 57 3 5 \n",
"514 16 0 0 15 0 12.67 37 3 5 \n",
"515 14 0 1 15 0 7.38 35 2 5 \n",
"516 16 0 0 10 0 15.56 32 3 5 \n",
"517 12 1 1 25 0 7.45 43 3 5 \n",
"518 14 0 1 12 0 6.25 32 3 5 \n",
"519 16 1 1 7 0 6.25 29 2 5 \n",
"520 17 0 0 7 1 9.37 30 3 5 \n",
"521 16 0 0 17 0 22.50 39 3 5 \n",
"522 16 0 0 10 1 7.50 32 3 5 \n",
"523 17 1 0 2 0 7.00 25 3 5 \n",
"524 9 1 1 34 1 5.75 49 1 5 \n",
"525 15 0 1 11 0 7.67 32 3 5 \n",
"526 15 0 0 10 0 12.50 31 3 5 \n",
"527 12 1 0 12 0 16.00 30 3 5 \n",
"528 16 0 1 6 1 11.79 28 3 5 \n",
"529 18 0 0 5 0 11.36 29 3 5 \n",
"530 12 0 1 33 0 6.10 51 1 5 \n",
"531 17 0 1 25 1 23.25 48 1 5 \n",
"532 12 1 0 13 1 19.88 31 3 5 \n",
"533 16 0 0 33 0 15.38 55 3 5 \n",
"\n",
" SECTOR MARR \n",
"0 1 1 \n",
"1 1 1 \n",
"2 1 0 \n",
"3 0 0 \n",
"4 0 1 \n",
"5 0 0 \n",
"6 0 0 \n",
"7 0 0 \n",
"8 1 1 \n",
"9 0 0 \n",
"10 0 1 \n",
"11 1 0 \n",
"12 0 1 \n",
"13 0 0 \n",
"14 0 1 \n",
"15 2 1 \n",
"16 0 1 \n",
"17 1 1 \n",
"18 0 0 \n",
"19 0 1 \n",
"20 1 1 \n",
"21 1 1 \n",
"22 0 0 \n",
"23 0 1 \n",
"24 0 1 \n",
"25 1 1 \n",
"26 1 1 \n",
"27 0 1 \n",
"28 1 1 \n",
"29 0 1 \n",
".. ... ... \n",
"504 0 0 \n",
"505 0 0 \n",
"506 0 1 \n",
"507 0 1 \n",
"508 0 1 \n",
"509 0 1 \n",
"510 0 1 \n",
"511 0 1 \n",
"512 0 1 \n",
"513 0 1 \n",
"514 0 1 \n",
"515 0 0 \n",
"516 0 0 \n",
"517 0 0 \n",
"518 0 1 \n",
"519 0 1 \n",
"520 0 1 \n",
"521 1 1 \n",
"522 0 1 \n",
"523 0 1 \n",
"524 0 1 \n",
"525 0 1 \n",
"526 0 0 \n",
"527 0 1 \n",
"528 0 0 \n",
"529 0 0 \n",
"530 0 1 \n",
"531 0 1 \n",
"532 0 1 \n",
"533 1 1 \n",
"\n",
"[534 rows x 11 columns]\n"
]
}
],
"source": [
"print data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1.4.1 配对图: 散点矩阵\n",
"\n",
"使用[seaborn.pairplot()](http://stanford.edu/~mwaskom/software/seaborn/generated/seaborn.pairplot.html#seaborn.pairplot)来显示散点矩阵我们可以很轻松的对连续变量之间的交互有一个直觉:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/matplotlib/collections.py:590: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
" if self._edgecolors == str('face'):\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAhcAAAIVCAYAAABiEd1DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import seaborn\n",
"seaborn.pairplot(data, vars=['WAGE', 'AGE', 'EDUCATION'], kind='reg') "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"可以用颜色来绘制类别变量:"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAIVCAYAAADS7vtxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDMJIsdQP1gEAAAAASUVORK5CYII=\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"seaborn.pairplot(data, vars=['WAGE', 'AGE', 'EDUCATION'], kind='reg', hue='SEX')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"---\n",
"\n",
"**看一下并感受一些matplotlib设置**\n",
"\n",
"Seaborn改变了matplotlib的默认图案以便获得更\"现代\"、更\"类似Excel\"的外观。它是通过import来实现的。重置默认设置可以使用:"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"from matplotlib import pyplot as plt\n",
"plt.rcdefaults()"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"要切换回seaborn设置, 或者更好理解seaborn中的样式, 见[seaborn文档中的相关部分](http://stanford.edu/~mwaskom/software/seaborn/tutorial/aesthetics.html)。\n",
"\n",
"---\n",
"\n",
"### 3.1.4.2. lmplot: 绘制一个单变量回归\n",
"\n",
"回归捕捉了一个变量与另一个变量的关系,例如薪水和教育,可以用[seaborn.lmplot()](http://stanford.edu/~mwaskom/software/seaborn/generated/seaborn.lmplot.html#seaborn.lmplot)来绘制:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFhCAYAAACh/xvXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"seaborn.lmplot(y='WAGE', x='EDUCATION', data=data) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"--- \n",
"\n",
"**稳健回归**\n",
"\n",
"在上图中,有一些数据点偏离了右侧的主要云,他们可能是异常值,对总体没有代表性,但是,推动了回归。\n",
"\n",
"要计算对异常值不敏感的回归,必须使用[稳健模型](https://en.wikipedia.org/wiki/Robust_statistics)。在seaborn的绘图函数中可以使用`robust=True`,或者在statsmodels用\"稳健线性回归\"`statsmodels.formula.api.rlm()`来替换OLS。\n",
"\n",
"## 3.1.5 交互作用检验\n",
"\n",
"![](http://www.scipy-lectures.org/_images/plot_wage_education_gender_1.png)\n",
"\n",
"是否教育对工资的提升在男性中比女性中更多?\n",
"\n",
"上图来自两个不同的拟合。我们需要公式化一个简单的模型来检验总体倾斜的差异。这通过\"交互作用\"来完成。"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: WAGE R-squared: 0.190\n",
"Model: OLS Adj. R-squared: 0.186\n",
"Method: Least Squares F-statistic: 41.50\n",
"Date: Thu, 19 Nov 2015 Prob (F-statistic): 4.24e-24\n",
"Time: 12:06:38 Log-Likelihood: -1575.0\n",
"No. Observations: 534 AIC: 3158.\n",
"Df Residuals: 530 BIC: 3175.\n",
"Df Model: 3 \n",
"Covariance Type: nonrobust \n",
"=========================================================================================\n",
" coef std err t P>|t| [95.0% Conf. Int.]\n",
"-----------------------------------------------------------------------------------------\n",
"Intercept 1.1046 1.314 0.841 0.401 -1.476 3.685\n",
"C(SEX)[T.1] -4.3704 2.085 -2.096 0.037 -8.466 -0.274\n",
"EDUCATION 0.6831 0.099 6.918 0.000 0.489 0.877\n",
"EDUCATION:C(SEX)[T.1] 0.1725 0.157 1.098 0.273 -0.136 0.481\n",
"==============================================================================\n",
"Omnibus: 208.151 Durbin-Watson: 1.863\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 1278.081\n",
"Skew: 1.587 Prob(JB): 2.94e-278\n",
"Kurtosis: 9.883 Cond. No. 170.\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
]
}
],
"source": [
"result = ols(formula='WAGE ~ EDUCATION + C(SEX) + EDUCATION * C(SEX)', data=data).fit() \n",
"print(result.summary())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"我们可以得出结论教育对男性的益处大于女性吗?\n",
"\n",
"---\n",
"**带回家的信息**\n",
"- 假设检验和p-值告诉你影响 / 差异的**显著性**\n",
"- **公式** (带有类别变量) 让你可以表达你数据中的丰富联系\n",
"- **可视化**数据和简单模型拟合很重要!\n",
"- **条件化** (添加可以解释所有或部分方差的因素) 在改变交互作用建模方面非常重要。\n",
"\n",
"---\n",
"\n",
"## 3.1.6 完整例子\n",
"\n",
"### [3.1.6.1 例子](http://www.scipy-lectures.org/packages/statistics/auto_examples/index.html)\n",
"#### [3.1.6.1.1 代码例子](http://www.scipy-lectures.org/packages/statistics/auto_examples/index.html#code-examples)\n",
"#### [3.1.6.1.2 课程练习的答案](http://www.scipy-lectures.org/packages/statistics/auto_examples/index.html#solutions-to-the-exercises-of-the-course)"
]
}
],
"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
}