Pyhon 裡面的Random與Pandas的Random

在看教學文章時，常常搞不清楚random.random , random.rand , random.randn 一堆看似類似但卻又不太一樣的函數。為了學的踏實點，就去了解了下: 

import random  # random just return one object import numpy as np  # use numpy.random  will return one or an array object   a = random.random() # return one object aa = random.random(10) # fail, because it only return one object   b = random.randint(10,20) # return an int object between 10~20   c = random.uniform(10,20) #return an object between 10~20   d = random.randrange(10,20,3) #return an int object in range 10~20 (such 10, 13,16,19)   e = random.choice(('pass','fail','null')) # random choice one object       A = np.random.random() # same as random.random()   AA = np.random.random(10) # 更新：不建議這樣用，因為np.random.random(10,4)的矩陣會回報錯 # can return an array   G = np.random.rand(10)   # 更新：np.random.rand(10,4)的矩陣仍可以過 所以用矩陣的話寫rand就好 # same as np.random.random(10)   F = np.random.randn() # return an normal distribution or Gaussian distribution object in random FF=  np.random.randn(10) # return an normal distribution or Gaussian distribution array in random       B = np.random.randint(10,20) # same as random.randint(10,20)   CC = np.random.uniform(10,20) # same as random.uniform(10,20)   DD = np.random.randrange(10,20,3) # fail, becauseno attribute randrange in np.random

總的來說，如果要出一個隨機數，就用random的就好。

如果需要用到array或矩陣的再引用np.random 免得搞混

這封郵件來自 Evernote。Evernote 是您專屬的工作空間，免費下載 Evernote