Julien Cohen-Adad
Julien Cohen-Adad
> @all-contributors please add jcohenadad for code @alexfoias looks like there is manual intervention needed-- can we have a fully-automated solution?
this is low priority
> Should I put them [here](https://github.com/spinalcordtoolbox/PAM50) or to a new repo ? Good question. I would suggest to integrate it in the PAM50 because users will want to e.g. warp...
> So we can follow skimage's pattern, but simply replace rng.normal() with numpy's rng.rayleigh() instead, then adjust the scale parameter of the distribution. I'm not sure that would work because...
ahah! indeed a dtype issue. When converting one of the image being subtracted to an UINT8: ~~~ sct_image -i sub-cardiff03_T2star_gmseg_crop.nii.gz -type uint8 -o sub-cardiff03_T2star_gmseg_crop_uint8.nii.gz sct_maths -i sub-cardiff03_T2star_seg_crop.nii.gz -sub sub-cardiff03_T2star_gmseg_crop_uint8.nii.gz -o...
> @jcohenadad This got closed via the gm-challenge commit -- do we want to leave this open for `sct_deepseg_gm`'s datatype issues and etc. ooops! that's a mistake-- i didn't want...
@joshuacwnewton this is a brilliant idea. Agreed with all the points you raised. Regarding the readability of some usages, eg [sct_register_multimodal](https://spinalcordtoolbox.com/en/stable/user_section/command-line.html#sct-register-multimodal), I am wondering if we could maybe come up...
> When you say "number of chars", you're referring to the width/length of each line, right? yup! > In our case, we have our custom SmartFormatter that we would adapt....
> To fix this, we could modify install_sct to download the source code, that way the script could be distributed and run by itself (e.g. install_sct-5.4.sh attached to the release)....
tagging @nekohayo so he is aware