additional trial by trial metrics - bias and plotting updates

[alexpiet]

• Updates plot.plot_foraging_session to include a wrapper function that takes NWBs, rather than passing in lists of choice and rewards. This is a quick fix for plot.plot_foraging_session does not operate on dataframes, but rather on lists of events.  #10, but not a great solution
• Adds bias to plot.plot_foraging_session.
• Cleans up spacing in plot.plot_foraging_session, removes unnecessary whitespace.
• resolves plot_foraging_session marks ignored trials on the "right" side, which gives the impression that the ignored trials should have been right choices or something #33
• Adds some session stats to plot_foraging_session, but more needed after metadata is cleaned up (finish migrating session stats to plot_foraging_session.py #34)
• Adds more metrics:
  • bias, depends on Updates to code to support online bias computation aind-dynamic-foraging-models#48

[hanhou]

• trial duration (stop-time - start-time) (@hanhou in process_nwb.py was this trial_start to trial_stop time, or gocue to gocue?)

I computed duration (and other lick stats) for each epoch defined here:
https://github.com/AllenNeuralDynamics/aind-foraging-behavior-bonsai-basic/blob/5bd0db90bf9182a969bcb3c52af77c8bb9a36f19/code/process_nwbs.py#L87-L93

[alexpiet]

@hanhou @rachelstephlee @hagikent @ZhixiaoSu This is ready for review.

[alexpiet]

Example of adding bias to plot_foraging_session.py

[hagikent]

just a comment on comments. I guess you are intentionally using df and df_trials?

[hagikent]

bias plot looks great.

---

can you make the ticks edge-aligned instead of centre-aligned?

e.g. for Right

Jeremiah wanted this change when we were working on posters/presentations and I thought it should be preemptively addressed here (instead of all of us get pushback individually).

[alexpiet]

@hagikent How does this look?
I also resolved #33 while I was at it

[alexpiet]

@hagikent @hanhou @rachelstephlee @ZhixiaoSu pinging about a review

[hagikent]

Looks good to me.

[rachelstephlee]

sorry, just got to this now.

i was testing this on a capsule. the figures are coming out all wonky to me in terms of dimensions. i've tried following your testing code. i think i'm missing something here....

capsule is here: (feel free to kick me out): https://codeocean.allenneuraldynamics.org/capsule/0613128/tree?cw=true

i've sent you my ipynb file.

can show you this during hackathon tomorrow also.

[alexpiet]

Merging after resolving issues. Thanks everyone