[高级编程技术作业-Week 14]IPython notebooks, Pandas, Statsmodels

Anscombe's quartet

Anscombe's quartet comprises of four datasets, and is rather famous. Why? You'll find out in this exercise.

	dataset	x	y
0	I	10	8.04
1	I	8	6.95
2	I	13	7.58
3	I	9	8.81
4	I	11	8.33

Part 1

For each of the four datasets...

• Compute the mean and variance of both x and y
• Compute the correlation coefficient between x and y
• Compute the linear regression line: y=β0+β1x+ϵy=β0+β1x+ϵ (hint: use statsmodels and look at the Statsmodels notebook)

Part 2

Using Seaborn, visualize all four datasets.

hint: use sns.FacetGrid combined with plt.scatter

Codes:

import random

import numpy as np
import scipy as sp
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

import statsmodels.api as sm
import statsmodels.formula.api as smf

sns.set_context("talk")

anascombe = pd.read_csv('anscombe.csv')
anascombe.head()

#1-1 Compute the mean and variance of both x and y 
x_mean = anascombe.groupby('dataset')['x'].mean()
print('mean of x', x_mean)
y_mean = anascombe.groupby('dataset')['y'].mean()
print('mean of y', y_mean)

x_var = anascombe.groupby('dataset')['x'].var()
print('variance of x', x_var)
y_var = anascombe.groupby('dataset')['y'].var()
print('variance of y', y_var)

#1-2 Compute the correlation coefficient between x and y
cor = anascombe.groupby("dataset")['x'].corr(anascombe['y'])
print('correlation coefficient between x and y', cor)

#1-3 Compute the linear regression line: y=beta_0+beta_1*x+epsilon
for i in range(0,4):
    X = anascombe[i*11:i*11+11]['x']
    Y = anascombe[i*11:i*11+11]['y']
    X = sm.add_constant(X)
    ols = sm.OLS(Y, X)
    reg_func = ols.fit()
    print('dataset '+str(i+1), "y = "+str(reg_func.params[0])+"+"+str(reg_func.params[1])+"x")

#2 Using Seaborn, visualize all four datasets
m = sns.FacetGrid(anascombe, col="dataset")    
m.map(plt.scatter, "x","y")
plt.show()