Huaqi (Harvey) Qiu

Still can't be reached :-(

How to make sure all torchsample dependancies are installed? I had to manually pip install every missing dependancy complained by the import call.

Just adding some details to @cherise215's answer to explain this implementation: - `n`-th order B-spline function can be computed by recursive convolution with `0`-th order B-spline function `n` times: ...

A small problem in my opinon is that the transformation model in this implementation doesn't seem to allow flexible combinations of kernel size and control point spacing (`stride=sigma` in transposed...

Not 100% sure that I understand you. Are you trying to use B-Spline transformation to generate random elastic deformation? If so, you can use the B-spline kernel transformation module. Filling...

sorry I must've missed the review request for #62. I can see this is a good workaround for my usecase, which is transforming the points which are in the un-normalised...

Thank you for raising the issue. `deepali.core` indeed doesn't contain these imports in module namespace but it has not been updated for 2 years. And the FFD example imports these...

as we discussed earlier, the `keepdim` flag should avoid this in the most recent version. Can you send a minimal example to reproduce the issue if this still doesn't work...