Question

I am reading from text files with the code below:

import numpy as np
my_data = np.genfromtxt(resultsDirectory+'/Points.txt', delimiter=' ')
PointX = my_data[:,5]
PointY = my_data[:,11]

My input files are typically like this –

ParamA : 0 ParamB : 7 ParamC : 0 ParamD : 1 Result : FAIL Time : 0 Epsilon : 0.5
ParamA : 0 ParamB : 11 ParamC : 0 ParamD : 1 Result : FAIL Time : 2 Epsilon : 0.5
ParamA : 0 ParamB : 7 ParamC : 0 ParamD : 3 Result : FAIL Time : 2 Epsilon : 0.25
ParamA : 0 ParamB : 13 ParamC : 0 ParamD : 1 Result : FAIL Time : 7 Epsilon : 0.25
ParamA : 0 ParamB : 7 ParamC : 0 ParamD : 4 Result : FAIL Time : 8 Epsilon : 0.125
ParamA : 0 ParamB : 8 ParamC : 0 ParamD : 2 Result : FAIL Time : 1 Epsilon : 0.125
ParamA : 0 ParamB : 8 ParamC : 0 ParamD : 3 Result : FAIL Time : 3 Epsilon : 0.125
ParamA : 0 ParamB : 8 ParamC : 0 ParamD : 4 Result : FAIL Time : 6 Epsilon : 0.125
ParamA : 0 ParamB : 9 ParamC : 0 ParamD : 2 Result : FAIL Time : 6 Epsilon : 0.125
ParamA : 0 ParamB : 10 ParamC : 0 ParamD : 2 Result : FAIL Time : 5 Epsilon : 0.125
ParamA : 0 ParamB : 14 ParamC : 0 ParamD : 1 Result : FAIL Time : 6 Epsilon : 0.125

When I extract PointX from this I get

PointX = [7 11 7 13 7 8 8 8 9 10 14]
PointY = [1 1 3 1 4 2 3 4 2 2 1]

Now, sometimes my text files contain only a single line or are even empty.

For example, if the text file has only one line, the my_data array is like this –

[  nan   nan   0.    nan   nan   7.    nan   nan   0.    nan   nan   1.
   nan   nan   nan   nan   nan  0.    nan   nan   0.5]

In this case, my_data.shape returns (21,).

However reading the array PointX or PointY gives me an error as IndexError: invalid index. I wanted PointX=[7] and PointY=[1]. Or if the text file is empty, it should be PointX=[] and PointY=[].

How exactly should I solve this problem? Also I need my PointX and PointY to be an array in order not to break the code which is dependent on it.

Thank you.

Answer 1

Unfortunately, genfromtxt returns a 1D array if given a file with only one line, and returns a 2D array if given more than one line. You could handle the discrepancy by reshaping:

import numpy as np
my_data = np.genfromtxt('data', delimiter=' ')
if my_data.ndim == 1:
    my_data = my_data.reshape(1, -1)
PointX = my_data[:,5]
PointY = my_data[:,11]
print(PointX)
print(PointY)

yields

[ 7.]
[ 1.]