general dataloader

[soumide1102]

@hickmank this should be ready for your review.

Don't know if Bryan can look at this PR as I am not being able to add him as a reviewer. Maybe he can see this once he's added to lanl github?

[bekaiser-LANL]

Hi @soumide1102 - we need to change "return start_img, end_img, Dt, self.channel_map" to "return start_img, self.channel_map, end_img, self.channel_map, Dt" to align with what @hickmank wants the method to return, i.e., it should pass the t=t_0 array, the t=t_0 map, the t=t_1 array, the t=t_1 map, and Dt.

[bekaiser-LANL]

need to change "return start_img, end_img, Dt, self.channel_map" to "return start_img, self.channel_map, end_img, self.channel_map, Dt" to align with what hickmank wants the method to return, i.e., it should pass the t=t_0 array, the t=t_0 map, the t=t_1 array, the t=t_1 map, and Dt. Also, the sequential_Dataset does not accommodate for changing channel maps as the sequence progresses. I think that's ok, we can do that later (if we ever need to)

[hickmank]

I don't think this will pass our linting right now. If you need help running that locally let me know.

1. The positions need to be passed through meshgrid so they are 2D arrays like the rest of the channels.
2. You need a 2D-array defining the detonator positions as a channel
3. If there are two parts with the same material those fields need to be added together into a single channel. Can you point me to this logic in this PR.
4. You need tests written for these new classes and functions.

[galegozi]

Hi @soumide1102 ,
I think your PR looks good.
I left a couple of comments regarding the coding style (not linting-related). These are just a couple of small things I thought that if they were fixed would make this PR better, thanks!

[hickmank]

@bekaiser-LANL @soumide1102 Looks like the dataset file needs to be linted. But, this PR is also changing a LOT of files. I think something is strange about how your branches are being merged. Can you take a look at it?

[soumide1102]

yeah, I'm trying to get rid of the extra commits. Looks like Bryan's copy of this branch picked up these already merged commits somehow.

[hickmank]

@soumide1102 @bekaiser-LANL This looks great. Functionally, very good. You just need to get it through the linter and write tests.

[hickmank]

@soumide1102 @bekaiser-LANL let's get this merged in. Let me know what you need help with.

[soumide1102]

@hickmank I'll push the tests in a few minutes. Just fixing the last linting complaints for the tests.

[soumide1102]

@hickmank @bekaiser-LANL I've pushed tests for the npz dataloader. If you're fine with the additions, we can merge this.

[hickmank]

Looks good to me! Yay!

[hickmank]

@bekaiser-LANL you need to re-review to unblock the merge. Assuming you're okay with the present state being merged.

[hickmank]

@bekaiser-LANL re-review please so we can merge this!