Python 如何使用文本文件创建 4 x 4 矩阵

Question

So I have a "txt" file and I want to create 4 by 4 matrices with it. I need to separate them from each "4 4". How can I do it? The "input.txt" file contains this:

4 4
55 55 55 56
66 66 66 67
77 77 77 78
88 88 88 89
4 4
1 2 3 4
2 2 2 2
3 3 3 3
4 4 4 4
4 4
11 12 13 14
22 24 24 25
33 34 35 36
44 45 46 47

Just like I said, I have to make individual matrices and the each matrix ends from "4 4". How can I separate them from there? I tried to use some methods but couldn't do it still. Thank you for spending your time to read this question.I tried this:

[i.lower().replace('4 4', '').split() for i in open('input.txt', 'r')]

The output is this:

[[],
 ['55', '55', '55', '56'],
 ['66', '66', '66', '67'],
 ['77', '77', '77', '78'],
 ['88', '88', '88', '89'],
 [],
 ['1', '2', '3', '4'],
 ['2', '2', '2', '2'],
 ['3', '3', '3', '3'],
 [],
 [],
 ['11', '12', '13', '14'],
 ['22', '24', '24', '25'],
 ['33', '34', '35', '36'],
 ['45', '46', '47']]

As you can see, it's not the exact thing that I want.

Answer 1

You can start by removing these lines that bother you:

>>> lines = [line for line in text_file.splitlines() if line != "4 4"]

Then create matrices with every chunk 4 successive rows:

>>> [np.fromstring("\n".join(lines[x:x+4]), sep=" ", dtype=int).reshape(4,4) 
     for x in range(0, len(lines), 4)]

Which output is a list of matrices of size (4, 4):

[array([[55, 55, 55, 56],
       [66, 66, 66, 67],
       [77, 77, 77, 78],
       [88, 88, 88, 89]]), 

 array([[1, 2, 3, 4],
       [2, 2, 2, 2],
       [3, 3, 3, 3],
       [4, 4, 4, 4]]), 

 array([[11, 12, 13, 14],
       [22, 24, 24, 25],
       [33, 34, 35, 36],
       [44, 45, 46, 47]])]

Answer 2

Just another example:

f = open('untitled.txt', 'r').read().splitlines()

# filter the '4 4' elements, and split on space
s = list(map(str.split, filter(lambda x: x != '4 4', f)))

# turn the list in to a list of matrices
m = [np.array(s[i:i+4]) for i in range(0, len(s), 4)]

Answer 3

This is a bit over-the-top but I'm learning, so I'd like to use this question to post a solution using a Context Manager . (This means you can use with , I'll give an example):

import numpy as np

class MyMatrixFile(object):
    def __init__(self, filename, mode):
        self.__filename = filename
        self.__mode = mode

    def __enter__(self):
        self.__open_file = open(self.__filename, self.__mode)
        self.__deserialise()
        return self.__myMatrixList

    def __exit__(self, *args):
        self.__open_file.close()

    def __deserialise(self):
        data = self.__open_file.read().splitlines()
        temp = list(map(str.split, filter(lambda x: x != '4 4', data)))
        self.__myMatrixList = [np.array(temp[i:i+4]) for i in range(0, len(temp), 4)]

If you have the above class in your python file, you can open your file types with the following:

with MyMatrixFile('test_matrix_file.txt', 'r') as matrix_list:
    # do stuff

Where matrix_list is your list of numpy.array objects. Using it like this just means the code processing the file is pushed in to the class (effectively hidden). Additionally, the file will close once you come out of the with indent.

This would be especially useful if you need to process lots of this type of file!