Error with ciftify_peaktable : "ValueError : must have equal en keys and value when setting an iterable".

[SamDaou]

Good afternoon Dr.Dickie,

I'm an M2 student of neuroscience. I performed a vertex-wise comparison of cortical thickness maps (fs_32k) with FSL_PALM, thus obtaining statistical results for each hemisphere. Thereafter, I merged bot files into a single cifti file dscalar.nii with wb_command 1.4.2.

As I had significant clusters with -log(p) values > 1.3, I decided to extract the significant clusters using ciftify_peaktable. When I set a max-threshold at 1.3 and let other parameters as defaults, I get a nifti file and a csv file containing information about my clusters. However, my two largest and most significant clusters (that can be seen in the nifti file) do not appear in the final csv file.

I suppose I should reduce the surface distance below 20 (default value), but when I do it, I get the following error :
"ValueError : must have equal en keys and value when setting an iterable".

Besides, no matter the values I choose for my parameters, I get deprecation warnings of the following type but they do not seem to prevent the program from working :
"deprecated from version: 2.1
will raise class <class 'nibble.deprecator.expireddeprecationerror'> as of version
4.0

Thank you in advance for your help,

DAOUDI Sami.

[edickie]

Hello, and thanks for trying ciftify! Do you get the same errors with ciftify_statclust_report? It should be replacing ciftify_peaktable functionality..

…

On Tue, May 25, 2021 at 11:17 AM SamDaou ***@***.***> wrote: Good afternoon Dr.Dickie, I'm an M2 student of neuroscience. I performed a vertex-wise comparison of cortical thickness maps (fs_32k) with FSL_PALM, thus obtaining statistical results for each hemisphere. Thereafter, I merged bot files into a single cifti file dscalar.nii with wb_command 1.4.2. As I had significant clusters with -log(p) values > 1.3, I decided to extract the significant clusters using ciftify_peaktable. When I set a max-threshold at 1.3 and let other parameters as defaults, I get a nifti file and a csv file containing information about my clusters. However, my two largest and most significant clusters (that can be seen in the nifti file) do not appear in the final csv file. I suppose I should reduce the surface distance below 20 (default value), but when I do it, I get the following error : "ValueError : must have equal en keys and value when setting an iterable". Besides, no matter the values I choose for my parameters, I get deprecation warnings of the following type but they do not seem to prevent the program from working : "deprecated from version: 2.1 will raise class <class 'nibble.deprecator.expireddeprecationerror'> as of version 4.0 Thank you in advance for your help, DAOUDI Sami. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#159>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADEXT5TPGGI3MBLAUVJY2D3TPO5R7ANCNFSM45PTFE2A> .

[SamDaou]

Good afternoon Dr.Dickie,

Thank you for you answer !

I tried using ciftify_statclust_report, but I'm getting an error (line 182 in the code) that I did not have with ciftify_peaktable, even when using the default parameters : AttributeError: 'Index' object has no attribute 'get_values'. Thereafter, only the nifti file is generated but not the csv file. I admit I do not know how to deal with that error ^^.

DAOUDI Sami.

[joebathelt]

I ran into the same issue. I think it might be because df.index.get_values() is deprecated in newer versions of pandas (>0.25, see pandas documentation). It could be replaced with df.index.to_numpy().

[joebathelt]

UPDATE: This did indeed solve it for me. I changed all instances of df.index.get_values() in ciftify_statclust_report.py and report.py to df.index.to_numpy. Then, ciftify_statclust_report ran and produced sensible outputs.