Rename common variables and functions

[tsalo]

Summary

There are a number of variables and functions that should be renamed across the package, in my mind, because they are confusing or uninformative.

A partial list

• catd --> data_cat
• tsoc --> data_optcom
• masksum --> adaptive_mask
• t2s --> t2s_limited
• s0 --> s0_limited
• t2ss --> t2s_ascending
• s0s --> s0_ascending
• t2sG --> t2s_full
• s0G --> s0_full
• dd --> data_dimred
• mmix --> mixing (sklearn convention)
• betas --> betas_optcom
• tedana.decomposition._utils.eimask() --> outlier_mask()
• tedana.stats.computefeats2() --> z_comp_maps()
• tedana.io.writeresults() --> tedana.io.write_results()
• tedana.io.writefeats() --> tedana.io.write_feats()
• tedana.io.load_data() --> tedana.io.reshape_data() (since it just reshaped input 4D data to 2D/3D)
• tedana.gscontrol.gscontrol_mmix() --> tedana.gscontrol.minimum_image_regression()
• tedana.gscontrol.gscontrol_raw() --> tedana.gscontrol.gscontrol()
• Finally, how do we feel generally about betas vs. parameter estimates vs. regression parameter, etc? I personally had enough classes with SPSS that I automatically interpret "beta" as "standardized parameter estimate", which is not how we use it throughout the package.

Next Steps

1. Discussion.
2. Large-scale renaming.

[dowdlelt]

I'm fine with betas as the label and it is nice and short. do we think it should include that these are ICA betas? As in, betas_ica_optcom? If the optcom was not used to generate the betas, would it be betas_sepechoes? or something like that?

Overall I am very much pro renaming to names that kind of mean something. Just reading this has helped me make sense of it.

[emdupre]

What do you think of ctab to comp_tbl ?

[rmarkello]

@emdupre: I think @handwerkerd had updated ctab to just comptable in his branch of the code, which would also work!

[tsalo]

ctab refers to the comptable file provided by the user though, while comptable is the DataFrame. I think there are times when we need separate variables. Same for the mixing matrix and T2* map (pending #408).

[tsalo]

I'd like this to be the final thing we address for 0.0.9.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions to tedana:tada: !

[tsalo]

Not stale.

[tsalo]

I propose we push this to 0.0.10.

[tsalo]

I'd like to start working on this in conjunction with #146 soon, given that there are relatively few active open PRs with major changes to the code. (My assumption is that the reports PR doesn't have much overlap with the files that need to be changed).

However, these changes will lead to major merge conflicts for any open pull requests, as well as conflicts for the proposed changes if new PRs are opened while this one is still unmerged, so I want to see what people think. I'd prefer to fast-track review of the proposed PR so that it doesn't get stuck and have conflicts accumulate.

Alternatively, I could break the changes into a bunch of separate PRs to keep the diffs manageable. What are folks' thoughts?

[jbteves]

My 2 cents:
I think we'll want to do one major PR and get it merged quickly. Every time we merge it'll make a downstream series of merge conflicts, so we want to minimize the number of merges. So we should do one merge and plan that after merge, we should coordinate to help people resolve conflicts on their forks.

[stale]

This issue has been automatically marked as stale because it has not had any activity in 90 days. It will be closed in 600 days if no further activity occurs. Thank you for your contributions to tedana:tada: !

[tsalo]

Happy to close this in favor of #919.