Question

任何人都知道是否有一种无需模块即可创建总面积表（SAT）的方法吗？我已经尝试过并工作：

import numpy as np

A = np.random.randint(0, 10, (3, 4))

print (A)
print(A.cumsum(axis=0).cumsum(axis=1))

但这是与numpy模块一起使用的。

Answer 1

In [99]: A = np.random.randint(0, 10, (3, 4))
In [100]: A
Out[100]: 
array([[1, 4, 0, 9],
       [5, 5, 3, 0],
       [3, 2, 1, 2]])
In [101]: A.cumsum(axis=0).cumsum(axis=1)
Out[101]: 
array([[ 1,  5,  5, 14],
       [ 6, 15, 18, 27],
       [ 9, 20, 24, 35]])

itertools有一个accumulate工具，其行为类似于cumsum：

https://docs.python.org/3/library/itertools.html#itertool-functions

In [102]: from itertools import accumulate
In [103]: Al = A.tolist()
In [105]: [list(accumulate(row)) for row in Al]
Out[105]: [[1, 5, 5, 14], [5, 10, 13, 13], [3, 5, 6, 8]]
In [106]: [list(accumulate(row)) for row in zip(*_)]
Out[106]: [[1, 6, 9], [5, 15, 20], [5, 18, 24], [14, 27, 35]]
In [107]: list(zip(*_))
Out[107]: [(1, 5, 5, 14), (6, 15, 18, 27), (9, 20, 24, 35)]

或打包成一个表达式：

In [109]: list(zip(*
     ...:    (accumulate(row) for row in zip(*
     ...:    (accumulate(row) for row in Al)))))
     ...:     
Out[109]: [(1, 5, 5, 14), (6, 15, 18, 27), (9, 20, 24, 35)]

如果您也不能使用itertools，请使用文档页面上提供的rough equivalent生成器。

或者具有这种小的累加功能：

def accum(alist):
    total = 0; res = []
    for a in alist:
        total += a
        res.append(total)
    return res

In [111]: list(zip(*
     ...:    (accum(row) for row in zip(*
     ...:    (accum(row) for row in Al)))))  
Out[111]: [(1, 5, 5, 14), (6, 15, 18, 27), (9, 20, 24, 35)]

不带模块python 3的汇总面积表

1 个答案: