Bug in make_local_connectivity_ones.py

[xychen-ION]

I have installed numpy, scipy and nibabel with python version 2.7 and I am trying to test the code on the test data. But I meet an error when I run make_local_connectivity_ones. The error shows below:

" sparse_w=append(sparse_w,R[nzndx],1) Traceback (most recent call last):

File "", line 1, in sparse_w=append(sparse_w,R[nzndx],1)

File "/home/fbip018/anaconda3/envs/cxy_image_3/lib/python2.7/site-packages/numpy/lib/function_base.py", line 5003, in append return concatenate((arr, values), axis=axis)

IndexError: axis 1 out of bounds [0, 1) "

This error happens in line 155 in make_local_connectivity_ones.py which is the code "sparse_w=append(sparse_w,R[nzndx],1)". I don't know how to deal with it and I am sure the numpy is installed.

Thank you

[mehrshadg]

append function in older versions of scipy (I suspect this package is written with older versions of scipy) ignored the wrong dimension passed as the third argument and when I read the code for make_local_connectivity_ones the matrix dimension is wrong. So, I think if you pass zero as the third argument the code must work: sparse_w=append(sparse_w,R[nzndx],0)

[xychen-ION]

Thanks mehrshadg! Yes, I have tried it and it works. But I am not sure if it is exactly what author means. Because the code in line 153 is "sparse_i=append(sparse_i,ondx1d[nzndx]-1,0)", it makes no sense that the mistake you said could happen. I also found more bugs when I tried other function with test data and no function can pass. Can you have all function passed? I have the same suspect that this package is written with older versions of scipy. I have tried an older version of scipy (0.16.1) but it still can not work.

[kelmok]

Hi @mehrshadg @xychen-ION @ccraddock,

I got the same issue running the code (scipy 1.2.1, numpy 1.16.4, python 2.7).

   if not os.path.isfile('rm_ones_connectivity.npy'): make_local_connectivity_ones(maskname, 'rm_ones_connectivity.npy')
  File "make_local_connectivity_ones.py", line 155, in make_local_connectivity_ones
    sparse_i=append(sparse_i,(ondx1d[nzndx-1]-1),0)
  File "..anaconda3/envs/example_env2/lib/python2.7/site-packages/numpy/lib/function_base.py", line 4694, in append
    return concatenate((arr, values), axis=axis)
AxisError: axis 1 is out of bounds for array of dimension 1

does anyone have an update, or know which numpy/scipy version is compatible?

Thanks

[caludia96]

I am getting the same AxisError. Please let me know of any updates. @ccraddock @kelmok

[AlexLZM]

Same error here. I changed 1 to 0 as @mehrshadg said and I have to change 'dtype' of variable 'a' to 'int' in binfile_parcellation.py to run all the scripts.

But unfortunately, all spatial correlation or time correlation output seems wrong. Tested with COBRE data subject 40046, n_cluster = 50.

But random clustering works fine.