R vs Python

0. 最近新學了 R 語言

• 把她拿來與 舊學 Python 並排比較學習。
• 拿她們來畫幾個機率與統計的分布。
• 發現她們各擅勝場。

1. R

• Normal, \(\chi^2\), T

quit
N= 1000
Z= rnorm(n= N)


k= 10

aL= c()
for (i in 1:k){
  aL= rbind(aL, rnorm(n= N)^2)}

X2= colSums(aL)

T= Z/(X2/k)^.5

B= 30

boxplot(Z, horizontal = TRUE); grid()

hist(Z, breaks = B); grid()

boxplot(X2, horizontal = TRUE); grid()

hist(X2, breaks = B); grid()

boxplot(T, horizontal = TRUE); grid()

hist(T, breaks = B); grid()

NA
NA

2. Python

• Normal, \(\chi^2\), T

reticulate::repl_python()

import numpy as np
import matplotlib.pyplot as pl
import pandas  as pd
import seaborn as sb
import scipy.stats as st


N= 1000
Z=   st.norm.rvs(size= N)

k= 10
X2=  sum([st.norm.rvs(size= N)**2 
          for i in range(k)]
          )
T=   Z/(X2)**.5

#%%
pl.cla()
ax= sb.boxplot(data=Z, orient='h')
ax.grid(True)
pl.show()

pl.cla()
ax= sb.histplot(Z, kde=True)
ax.grid(True)
pl.show()

pl.cla()
ax= sb.boxplot(data=X2, orient='h')
ax.grid(True)
pl.show()

pl.cla()
ax= sb.histplot(X2, kde=True)
ax.grid(True)
pl.show()

pl.cla()
ax= sb.boxplot(data=T, orient='h')
ax.grid(True)
pl.show()

pl.cla()
ax= sb.histplot(T, kde=True)
ax.grid(True)
pl.show()