Add regression tests for T2*/S0 estimation and optimal combination

[tsalo]

Summary

Per discussion with @afni-rickr and @emdupre, we may want regression tests for our more stable steps, like T2*/S0 estimation with log-linear and optimal combination with the t2s method.

Additional Detail

We will need to tune our tolerances for the regression tests.

Next Steps

• [ ] Add CI job to download ground truth data from OSF
• [ ] Write tests that compare generated files to ground truth files with nib-diff (see #59)
• [ ] Choose tolerance levels

[stale[bot]]

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: !

[jbteves]

Boo stale-bot, we're doing our best.

[stale[bot]]

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: !

[stale[bot]]

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: !