text file parsing and importing permuted result

Question

I know the cvs module provides a very nice way to parse text files in python. I have a file that has the following columns:

x y z b

Here, xyzb are floating point numbers. However, I am faced with the possibility that the order right be permuted in any possible way. For example, it may be that the file comes as

b y x z

Now, I would like to read the data into a numpy array which should store the data in the bxyz format. So, the reading function should be something like:

def read_file(file_path, inputformat='xyzb', delimiter=' '):

Now, I could look at each line and have a big if/else kind of logic where for each of the four positions, I would query the format character and then insert it into the corresponding position in the array. However, I was wondering if there is more elegant way to take care of this permutation?

Answer 1

I would suggest loading the data using numpy.loadtxt (see documentation ), then array-slicing to correct the column order.

Edit: actually, you can use the usecols parameter to directly specify the desired column order, like:

import numpy as np

def read_file(file_path, input_format="xyzb", desired_format="bxyz", delimiter=" "):
    usecols = [input_format.index(col) for col in desired_format]
    with open(file_path, "r") as inf:
        return np.loadtxt(inf, usecols=usecols, delimiter=delimiter)