Dl wrapper + DL testing #109
Conversation
I built a wrapper in which DL algorithm train at initialisation
- If self-supervised, one can give data to the algorithm to train from.
- If supervised, or if no data is given, data is simulated
- For IVIM-NET this happens within the wrapper
- For Super IVIM DC this happens in the package (as I did not see an option to give training data)
Then testing occurs. For speed, I give all testing data in 1 go. Also, as deep learning is known to predominantly work better than LSQ in noisy data, and actually do a poor job in noise-less data, I made a second DL-specific dataset which contains much more noise. Also, I made DL-specific boundaries for passing unit testing.
|
Hey reviewers, Before you review some points: Ivan: For DL, it is often easier to throw through all data in 1 go as a 2D array (voxels, b-values). I think I now throw this though the single-voxel fit and put an if-statement. Maybe we can come up with a way to deal with this properly within the wrapper? All: I have added an additional data-set for AI-algorithms as these algorithms typically characterise themselves by working well in noisy data, but usually perform poorly in noise-less data. I also added broader margins. We can discuss whether this is the best option. All: |
… errors after merge :)
There was a problem hiding this comment.
A couple of comments regarding the changes to the wrapper. I also checked the standardized implementations and I don't see anything weird. Although we could think about adding supported_parallel_cpu and supported_gpu attributes now that we are getting more algorithms that do support it.
- Now use ivim_fit_full_volume instead - Skipped
|
I think this is running now and allows testing of AI algorithms. Happy to merge! |
| Returns: | ||
| _type_: _description_ | ||
| """ | ||
| signals = self.osipi_reshape_to_voxelwise(signals) |
There was a problem hiding this comment.
Beware that there is still a call to osipi_reshape_to_voxelwise here. I don't think this shows up in unit tests as we don't test this full_volume function there
Describe the changes you have made in this PR
A clear and concise description of what you want to happen
Link this PR to an issue [optional]
Fixes #ISSUE-NUMBER
Checklist