Skip to content

docs: run-log encoding options for course logs (turning + forward)#141

Open
mbreiser wants to merge 2 commits into
mainfrom
docs/run-log-encoding-options
Open

docs: run-log encoding options for course logs (turning + forward)#141
mbreiser wants to merge 2 commits into
mainfrom
docs/run-log-encoding-options

Conversation

@mbreiser

@mbreiser mbreiser commented Jul 4, 2026

Copy link
Copy Markdown
Contributor

A proposal doc to review with Frank on how to log FicTrac data for the course run-logs — balancing behavioural coverage vs. file size.

docs/development/run-log-encoding-options.md covers:

  • The 4 curated fields for turning + forward locomotion (delta-rotation lab side/fwd/yaw = cols 6–8, plus integrated heading = col 17), with the FicTrac axis convention + a bench sanity-check.
  • An encoding-options table from full-25-columns-as-JSON down to compact CSV, with sizes modeled on a 15-minute run @ 50 Hz (~45,000 frames):
    • Full 25 as JSON ≈ 16 MB → curated CSV (relative-ms, rounded) ≈ 1.8 MB (−89 %), still human-readable and self-describing; gzip optional (~0.5 MB).
  • A recommended hybrid scheme (JSON event rows + bare CSV frame rows + a frame_schema line) with a concrete example.
  • Open questions for Frank (field set, frame rate, archival of all-25, gzip acceptability, float precision / seq).

Implementation is tracked in #140 — this PR is the shareable discussion doc; not required to merge as-is.

mbreiser and others added 2 commits July 4, 2026 15:02
…zes @ 15-min)

Proposal doc (for Frank / #140): the 4 curated FicTrac fields for turning +
forward locomotion (delta-rot lab side/fwd/yaw cols 6-8 + integrated heading
col 17), and an encoding-options table from full-25-JSON down to compact CSV,
with file sizes modeled on a 15-minute run @ 50 Hz (45k frames). Recommends the
hybrid CSV-frames + JSON-events scheme (relative-ms, rounded floats) at ~1.8 MB.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…ted fields

Add an explicit 'identity + timing' table (ms/seq/index) and put seq back into
the recommended CSV schema; a full frame row is ms,seq,index,side,fwd,yaw,hd.
Note seq is the only droppable field (~8 B). Reconcile the size note (~2.2 MB
with seq / ~1.8 MB without over 15 min).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@mbreiser mbreiser requested a review from floesche July 4, 2026 19:52
@mbreiser

mbreiser commented Jul 4, 2026

Copy link
Copy Markdown
Contributor Author

@floesche Please take a look at the proposal and let me know if you agree with the recommendation for updating the data logging, or if you prefer storing different fields in a different format.

@floesche

floesche commented Jul 4, 2026

Copy link
Copy Markdown
Member

I made some suggestions in #143. I think the minimal recording is sufficient: FicTrac is pretty efficient in recording the data, but that would require a post-hoc step to combine the data.
If you just want to have something that is easy to plot and interpret, I would go for columns 15…18 instead. If the alignment is perfect, that shouldn't make much of a difference. So far, I don't think we have a well calibrated fictrac config just yet…

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants