NumPy Input/Output

There are two useful input functions that make it easy to read ascii text (normal text file) data within NumPy, loadtxt() and genfromtxt(). The two functions are very similar and the only difference comes in if you have missing data, in which case you would need to use genfromtxt(). The best cases to use these functions is with a file that contains tabular style data that is all separated by a delimiter (i.e., white space, comma). If there is not any a common delimiter, there are other methods that can be employed (Python I/O).

Assuming you have a file that is appropriate to use, there are a number of keyword arguments that can be added to either function to aid in getting the data input in the correct format, to one or more variables.

input.txt:

34 56 78 91
12 23 33 51
20 39 48 57

Code:

import numpy as np

filename = 'input.txt'
data = np.loadtxt(filename)
print(data)

Output:

[[ 34.  56.  78.  91.]
 [ 12.  23.  33.  51.]
 [ 20.  39.  48.  57.]]

The data is input into the program by the loadtxt() function from NumPy as a 2D array of real values and has the same row/column configuration.

Saving data to a text file is also pretty simple (savetxt()) and just involves naming the output file and giving the function the 1D or 2D array that you wish to save to the file.

np.savetxt('output_file.txt', data)

In the above example, instead of specifying the file name as a separate variable, it was just placed as a string where it needed to go in the function.