Use dev cfg for biasing, biassteps, bgmap. Remove fit_tmin kwarg

[dpdutcher]

This PR does several things related to biasing and bias steps:

• Adds dev-cfg defaults for bias_dets, both to_rfrac and to_rfrac_range
• Incorporates Adds dev-cfg defaults for bias_steps and take_bgmap #404 , which adds dev-cfg defaults for take_bias_steps and take_bgmap in the same manner as done for take_iv
• Makes fitting for the time constant optional, default True.
• Eliminates the kwarg fit_tmin in favor of determining the tmin from the data itself on a per-channel basis.

Draft PR until I've been able to test all aspects of these changes.

[dpdutcher]

For justification of the fit_tmin change, see https://simonsobs.atlassian.net/wiki/spaces/SOPS/pages/1379729421/Improving+time+constant+fitting+from+bias+steps

[dpdutcher]

I've tested bg_map, biasing, and bias_steps on an LF UFM in the Princeton testbed. I've tested analysis of historical bias step data from MF and UHF lab data, and UHF in-field data with and without a spinning HWP.
All worked as expected.

This will change how time constants are computed for all frequencies, but it's strictly for the better. All previous time constants should be recomputed.

[dpdutcher]

I'm noticing now that the way I've implemented rfrac(range) defaults won't work with the way the pysmurf_controller and and sorunlib is written, and that those will also need to be changed.
For example, having defaults for both 'rfrac' and 'rfrac_range' means there's a not a single parameter that's either a tuple or int, deciding between bias_to_rfrac or bias_to_rfrac_range. I'm not sure where that choice should be made.

[msilvafe]

A couple comments and questions along with a response to your question about rfrac vs rfrac_range inline.

[dpdutcher]

I've implemented and tested my changes of merging bias_to_rfrac and bias_to_rfrac_range. It pulls from the dev cfg correctly, and the user can pass confused calls like bias_to_rfrac(S,cfg, rfrac=(0.3, 0.6)) or bias_to_rfrac_range(S, cfg, rfrac=0.5) and those still work, too.

[msilvafe]

Thanks for testing and addressing comment.