porespy

porespy copied to clipboard

Make sure they are all consistent.

jgostick

This was more trouble that it was worth.

jgostick

`sierpinski_foam2` has a more sensible api where you actually specify the image size you want, and it crops accordingly. I think it would be smart to remove the original one...

jgostick

All tqdm instances should include a description.

1

The `tqdm` function accepts the `desc` keyword, which will then include a text description of what the progress bar is doing. We should probably fetch the function name where `tqdm`...

jgostick

This would be useful for making movies which are properly scaled in time.

jgostick

AI-assisted diffusive conductance model is failing, possibly a `tensorflow` API change

2

@jgostick Maybe the UG student who's working on Niloo's work but on hydraulic conductance can take a look? PS. I've disabled the tests for now, until this is fixed.

ma-sadeghi

Should funcs return a normal dict instead of a "results" object?

1

The results object has attributes attached like `result.im_blah`, while a `dict` would be `result['im_blah']`. The dict is a bit more 'normal' but lacks other custom functionality, like printability.

jgostick

It would be more convenient if all generators features would be either relative to the lengthscale of the image (`shape.mean()`) or to the size of the image (`np.prod(shape)`). For example,...

amirDahari1

`two_point_correlation_fft` is only calculated for one orthant

1

`two_point_correlation` calculates the average of the TPC of only one orthant (quadrant 2D/ octant 3D), which might be different than the average over all vector directions.

amirDahari1

Finalizing tortuosity_bt and rev functions

1

jgostick