Reading Matrix from file

Question

I have a txt file consisting some numbers with space and I want to make it as three 4*4 matrixes in python. Each matrix is also divided with two symbols in the text file. The format of the txt file is like this:

1 1
0 0 0 0
0 0 0 0
0 0 0 0 
0 0 0 0
1 1
0 0 0 0
0 0 0 0
0 0 0 0 
0 0 0 0
1 1
0 0 0 0
0 0 0 0 
0 0 0 0
0 0 0 0

My code is now like this but it is not showing the output I want.

file = open('inputs.txt','r')
a=[]
for line in file.readlines():
    a.append( [ int (x) for x in line.split('1 1') ] )

Can you help me with that?

Hyphenated

What is the meaning of lines with 1 1, do they separate matrixes? – Duello

The format of the output needs to be a three 4*4 array(matrixes) using numpy. 1 1 is the separation of the matrixes – Hyphenated

you should rather read all - read() and later split it with split('1 1') to create 4 strings with matrixes. And later you would have to split these strings using "\n" and space. – Granophyre

maybe first ise print() to see what you get in line and line.split('1 1') and x (before you use int(). It is called "print debuging" and it helps to see what code is doing. – Granophyre

Storing the whole file into memory is not recommended. It might be terabytes... – Jerad

Answer 1

A good old pure python algorithm (assuming matrices can hold string values, otherwise, convert as required):

file = open("inputs.txt",'r')
matrices=[]
m=[]
for line in file:
   if line=="1 1\n": 
      if len(m)>0: matrices.append(m)
      m=[]
   else:
      m.append(line.strip().split(' '))
if len(m)>0: matrices.append(m)
print(matrices)
# [[['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0']], 
#  [['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0']], 
#  [['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0'], ['0', '0', '0', '0']]]

Jerad

How can I make it three 4*4 matrixes? For example writing the first row in a 4*4 matrix? I want them to be shown not right to each other but like 4 different lines under each other – Hyphenated

There are no such things as matrixes in python. Only lists. You might want to look at numpy for examples, but I don't know exactly what you need. BTW screen output is independent of storage in memory. – Jerad

Answer 2

One option is to use groupby:

from itertools import groupby

matrices = []

with open('inputs.txt', 'r') as f:
    for separator, lines in groupby(f, lambda line: line.strip() == '1 1'):
        if not separator:
            matrices.append([[int(x) for x in line.split()] for line in lines])

print(matrices)
# [[[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]],
#  [[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]],
#  [[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]]

Reef

How can I make it three 4*4 matrixes? For example writing the first row in a 4*4 matrix? I want them to be shown not right to each other but like 4 different lines under each other – Hyphenated

Answer 3

Assuming your file is not humongous, it is numerical and 4x4, the easiest method is:

• read all the file
• split it in blocks with the 1 1\n separator
• discard the separator items (if block)
• convert block to vector with split
• reshape each vector to 4x4
• make it integer

In a single line: matrixes = [np.reshape(np.array(block.split()),(4,4)).astype(int) for block in open('inputs.txt').read().split('1 1\n') if block]

Caveat: if a matrix reads x x 1 1 in one of its rows, it will be considered a split regardless. Using a value that could be used in the matrix is not a good idea.

That could be prevented splitting on \n1 1\n, and removing by hand the first 4 characters (1 1\n). Also this implementation may be more efficient, flattening everything and then reshaping:

dd = open('inputs.txt').read()[4:]
nmats = dd.count('\n1 1\n') +1
matrixes = np.reshape(np.array(dd.replace('\n1 1\n',' ').split()).astype(int),(nmats,4,4))

This last option returns it as a single 3D matrix:

>>> matrixes
array([[[0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0]],

       [[0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0]]])
>>> matrixes[0]
array([[0, 0, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 0]])

Sonata