Clean up regress

[ejolly]

It would be great to add a few options to run with Brain_Data.regress() that users don't often do themselves. Not defaults, just kwargs:

• VIF check on passed Design_Matrix
• Scaling to interpret betas as % signal change
• Adjusting SE epsilon factor to prevent divide-by-zeros or using np.lstsqrs

[ejolly]

v0.6.0 Status

These regress() enhancements are still relevant for a future update:

• VIF check on passed Design_Matrix
• Scaling to interpret betas as % signal change
• Adjusting SE epsilon factor

The core regress functionality remains stable in v0.6.0. These are quality-of-life improvements that could be added as optional kwargs.