CODEWITHSUNDEEP
pythonmatlabnumpyrangemultidimensional-array
Programmer
Programmer

Reputation: 8727

Numpy / Scipy : why it is throwing error

I am trying to convert a MATLAB code into Python.

My MATLAB Code:

ASE_lamda1=1000e-9;        
ASE_lamda2=1100e-9;        
del_lamda= 2e-9;           
ASE_lamda = (ASE_lamda1:del_lamda: ASE_lamda2)';

Below is what I am trying as eqv. Python code:

#!/usr/bin/python

import numpy as np

ASE_lamda1 = 9.9999999999999995e-07
ASE_lamda2 = 1100e-9
del_lamda = 2e-9
ASE_lamda = np.transpose(np.arange[ASE_lamda1:del_lamda:ASE_lamda2])

But I am getting the below error:

Traceback (most recent call last):
  File "tasks.py", line 22, in <module>
    ASE_lamda = np.transpose(np.arange[ASE_lamda1:del_lamda:ASE_lamda2])
TypeError: 'builtin_function_or_method' object has no attribute '__getitem__'

I am not sure of this error - what it means as I do not have much expertise over Python / Numpy / Scipy.

Upvotes: 3

Views: 102

Answers (1)

NPE
NPE

Reputation: 500933

The

np.arange[ASE_lamda1:del_lamda:ASE_lamda2]

should be

np.arange(ASE_lamda1, ASE_lamda2, del_lamda)

This returns

array([  1.00000000e-06,   1.00200000e-06,   1.00400000e-06,
         1.00600000e-06,   1.00800000e-06,   1.01000000e-06,
         ...
         1.09000000e-06,   1.09200000e-06,   1.09400000e-06,
         1.09600000e-06,   1.09800000e-06,   1.10000000e-06])

This is a 1D array, so transposing it is a no-op. You may or may not need to reshape it to 2D depending on what you're going to do with it. An easy way reshape the array to 2D is using slicing and numpy.newaxis:

In [54]: ASE_lamda[:, np.newaxis]
Out[54]: 
array([[  1.00000000e-06],
       [  1.00200000e-06],
       ...
       [  1.09800000e-06],
       [  1.10000000e-06]])

In [55]: ASE_lamda[np.newaxis, :]
Out[55]: 
array([[  1.00000000e-06,   1.00200000e-06,   1.00400000e-06,
          1.00600000e-06,   1.00800000e-06,   1.01000000e-06,
          ...
          1.09000000e-06,   1.09200000e-06,   1.09400000e-06,
          1.09600000e-06,   1.09800000e-06,   1.10000000e-06]])

If you're moving to NumPy from MATLAB, take a look at NumPy for Matlab Users.

Upvotes: 2

Related Questions