Open thewtex opened 1 year ago
Perhaps it makes more sense to get the python included into the Pypi package as was done here:
https://discourse.itk.org/t/include-additional-files-in-itk-external-module-python-wheels/5794/7
?
On Tue, Sep 19, 2023 at 10:52 AM Matt McCormick @.***> wrote:
Hi @aylward https://github.com/aylward @brad-t-moore https://github.com/brad-t-moore @dzenanz https://github.com/dzenanz ,
Loving this project.
In order to facilitate code re-use and maintenance, I am wondering what you think of splitting the pure python modules in wrapping/ into a separate package, perhaps itk-argus-py, that would depend on the C++-wrapped itk-argus package. The package could be configured easily in a subdirectory using hatch https://hatch.pypa.io/latest/?
— Reply to this email directly, view it on GitHub https://github.com/KitwareMedical/itkARGUS/issues/6, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACEJL6MY77OOI3J2E5XJEDX3GWSZANCNFSM6AAAAAA46OHKKI . You are receiving this because you were mentioned.Message ID: @.***>
-- Stephen R. Aylward, Ph.D. Chair, MONAI Advisory Board Senior Director, Strategic Initiatives, Kitware
@thewtex what do you think would likely be less effort? Hatch or RTK's trick?
I'd rather maintain only one package and have a path for continuously improving itkARGUS, rather than create a fracture based on the underlying code.
Even if it is a bit harder, it seems like the right thing is to keep them combined.
s
On Tue, Sep 19, 2023 at 11:44 AM Dženan Zukić @.***> wrote:
@thewtex https://github.com/thewtex what do you think would likely be less effort? Hatch or RTK's trick?
— Reply to this email directly, view it on GitHub https://github.com/KitwareMedical/itkARGUS/issues/6#issuecomment-1725920735, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACEJL26F22WU7V3ZGZMSPTX3G4VZANCNFSM6AAAAAA46OHKKI . You are receiving this because you were mentioned.Message ID: @.***>
-- Stephen R. Aylward, Ph.D. Chair, MONAI Advisory Board Senior Director, Strategic Initiatives, Kitware
Notes from discussion with @aylward and his suggestion:
A good path forward could be to port the deep learning ITK C++ scan conversion code into itk-ultrasound, then publish this as a pip-installable pure python package (hatch is the easy-best-practice way), highlighting the MONAI transform for ultrasound time series and related code.
Both C++ classes here seem specific to ARGUS, and probably not interesting in ITKUltrasound. Otherwise, migrating them into ITKUltrasound seems easy.
Hi @aylward @brad-t-moore @dzenanz ,
Loving this project.
In order to facilitate code re-use and maintenance, I am wondering what you think of splitting the pure python modules in wrapping/ into a separate package, perhaps
itk-argus-py
, that would depend on the C++-wrappeditk-argus
package. The package could be configured easily in a subdirectory using hatch?