Python求中位數(shù)

瑞安3980

[[1,2,3,4],
[5,6,7,8],
[9,10,11,12],
[13,14,15,16],
[17,18,19,20]]這組列表怎么求內(nèi)置列表每列的中位數(shù)，注意是求1,5,9,13,17和2,6,10,14,18等列的中位數(shù)，怎么求其中位數(shù)？

gxyzwangyi

本人新手，隨手寫了一段不那么pythonic的代碼，不過(guò)結(jié)果應(yīng)該是對(duì)的

1. 
   
2. a=[[1,2,3,4],[5,6,7,8],[9,10,11,12],[13,14,15,16],[17,18,19,20]]
   
3. b=[[] for i in range(len(a[0]))]
   
4. for i in range(len(a[0])):
   
5.     for j in range(len(a)):
   
6.         b[i].append(a[j][i])
   
7. for i in range(len(b)):
   
8.     b[i]=sorted(b[i])[int(len(b[i])/2)]
   

復(fù)制代碼

icymirror

回復(fù) 1# 瑞安3980
給寫了段code，你以后可以只更換傳入的函數(shù)來(lái)對(duì)于list in list的數(shù)值的計(jì)算了。
比如：你自己再寫個(gè)stdev等等。

1. # do not apply to empty list, it will result exception
   
2. mean = lambda x: sum(x) * 1.0 / len(x)
   
3. median = lambda x: sorted(x)[len(x) / 2] if len(x) % 2 == 1 else (sorted(x)[len(x) / 2 - 1] + sorted(x)[len(x) / 2]) / 2.0
   
4. 
   
5. def MatrixCalc(data, function, byRow=True):
   
6.     """
   
7.     """
   
8.     if byRow: # By Column
   
9.         return map(function, data)
   
10.     else: # By Column
    
11.         return map(function, map(list, zip(*data)))
    
12. 
    
13. def main():
    
14.     data = [[1, 2, 3, 10], [2, 3, 4, 5], [3, 3, 9, 100]]
    
15.     print MatrixCalc(data, mean, True)
    
16.     print MatrixCalc(data, median, True)
    
17.     print MatrixCalc(data, mean, False)
    
18.     print MatrixCalc(data, median, False)
    
19. 
    
20. if __name__ == "__main__":
    
21.     main()

復(fù)制代碼

inpool

1. a=[[1,2,3,4],[5,6,7,8],[9,10,11,12],[13,14,15,16],[17,18,19,20]]
   
2. 
   
3. def zhong(list_):
   
4.     result = []
   
5.     for i in zip(*list_):
   
6.         nums = sorted(i)
   
7.         length = len(x)/2
   
8.         if len(x) % 2 == 0:
   
9.             result.append(sum(nums[length-1:length+1]) / 2.0)
   
10.         else:
    
11.             result.append(nums[length])
    
12.     return result
    
13. 
    
14. print zhong(a)

復(fù)制代碼

tianzhizhengdao

1. import numpy as np
   
2. 
   
3. a = [[1,2,3,4],[5,6,7,8],[9,10,11,12],[13,14,15,16],[17,18,19,20]]
   
4. b = [values for sublist in a for values in sublist]
   
5. 
   
6. # Decent way to flat list
   
7. # import itertools
   
8. # b = list(itertools.chain.from_iterable(a))
   
9. 
   
10. # Simplest way
    
11. # b = sum(a, [])
    
12. 
    
13. median = [np.median(b[i::4]) for i in range(4)]

復(fù)制代碼