numpy 1.12.0 rejects floats as reshape arguments
gdevenyi opened this issue · 8 comments
This section of code
https://github.com/SebastianoF/bruker2nifti/blob/master/bruker2nifti/_getters.py#L161-L166
Dies on numpy 1.12.0 with the following error:
/home/cic/desgab/Desktop/bruker2nifti_qa-master/raw/McGill_Orientation/a20130329_APM_DEV_Orient.j71
started:
Converting experiment 1:
pdata/
1/
d3proc
id
meta
roi
2dseq
visu_pars
procs
reco
Traceback (most recent call last):
File "test_b2n.py", line 9, in <module>
bru.convert()
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 185, in convert
nifti_file_name=scan_name)
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 137, in convert_scan
get_reco=self.get_reco,
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_cores.py", line 150, in scan2struct
consider_subject_position=consider_subject_position
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_getters.py", line 166, in nifti_getter
vol_data = img_data_vol.reshape(vol_pre_shape, order='F')
TypeError: 'numpy.float64' object cannot be interpreted as an integer
Thanks! Maybe solved on the dev branch.
Can you please check if it works on your side, as I could not reproduce the error with my current dataset.
Still bugged, now with a reshape problem
/home/cic/desgab/Desktop/bruker2nifti_qa-master/raw/McGill_Orientation/a20130329_APM_DEV_Orient.j71
started:
Converting experiment 1:
pdata/
1/
d3proc
id
meta
roi
2dseq
visu_pars
procs
reco
Traceback (most recent call last):
File "test_b2n.py", line 9, in <module>
bru.convert()
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 185, in convert
nifti_file_name=scan_name)
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 137, in convert_scan
get_reco=self.get_reco,
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_cores.py", line 150, in scan2struct
consider_subject_position=consider_subject_position
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_getters.py", line 167, in nifti_getter
vol_data = img_data_vol.reshape(vol_pre_shape, order='F')
ValueError: cannot reshape array of size 66 into shape (256,256,15,0)
Uhm... It seems that the information in the Bruker parameters files does not match the dimension of the array. Try to copy your data in a new place, erase folder 1 and see if the other experiments are converted correctly. So I will know how much do I have to worry :-) ! Let me know.
BTW, this is the test data from the repo....
Study conversion
/home/cic/desgab/Desktop/bruker2nifti_qa-master/raw/McGill_Orientation/a20130329_APM_DEV_Orient.j71
started:
Converting experiment 2:
pdata/
1/
d3proc
meta
2dseq
procs
visu_pars
roi
id
reco
Traceback (most recent call last):
File "test_b2n.py", line 9, in <module>
bru.convert()
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 185, in convert
nifti_file_name=scan_name)
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/converter.py", line 137, in convert_scan
get_reco=self.get_reco,
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_cores.py", line 150, in scan2struct
consider_subject_position=consider_subject_position
File "/home/cic/desgab/Desktop/bruker2nifti/bruker2nifti/_getters.py", line 167, in nifti_getter
vol_data = img_data_vol.reshape(vol_pre_shape, order='F')
ValueError: cannot reshape array of size 65 into shape (64,64,32,0)
Hi, sorry I had time to look into it only now.
The odd thing is, I could not reproduce the error with the same dataset (bruker2nifti_qa-master downloaded 30 Mar 2015).
With python3, numpy==1.12.0, bruker2nifti installed in development mode, on a mac the error I get is:
File "/Applications/python_virtualenvs/clean3/lib/python3.6/site-packages/pkg_resources/__init__.py", line 774, in resolve
raise DistributionNotFound(req, requirers)
pkg_resources.DistributionNotFound: The 'numpy>=1.12.1' distribution was not found and is required by bruker2nifti
With numpy==1.12.1 no errors are raised.
Idem if using python2.
I am unable to reproduce my user's errors in a clean environment, so I'm considering the original bug fixed and closed. Thanks.
Followup: error due to user not having git-lfs when getting a copy of the QA repo.