Vertical slice

IBL Visual Decision

First visual 2AFC slice with OpenAlyx provenance, canonical trials, and fitted descriptive psychometrics for one public IBL session.

Run in your browser

IBL Visual Decision · psychometric by prior block

Group canonical trials by prior block (probabilityLeft) and signed contrast, then compute P(right) and trial counts per cell. Useful as a starting point for fitting prior-conditioned psychometrics.

Pyodide + scipy run locally; first load downloads ~12 MB.

IBL Visual Decision · psychometric by prior block

Pyodide 0.26.4

Group canonical trials by prior block (probabilityLeft) and signed contrast, then compute P(right) and trial counts per cell. Useful as a starting point for fitting prior-conditioned psychometrics.

Python (df is a pandas DataFrame loaded from the trials CSV; ⌘/Ctrl-Enter to run)
Click run to start

Runs entirely in your browser. First click downloads ~10 MB of Pyodide and pandas; subsequent runs reuse the cached runtime.

Findings extracted

Curves attached to this slice, on shared signed-evidence axes.

Metrics

Trials 569
Response trials 537
No-response trials 32
Prior blocks 3

Slice profile

Operational variables for cross-slice comparison.

Species
mouse
Modality
visual
Stimulus metric
signed contrast
Evidence type
static
Choice type
2afc
Response modality
wheel
Source data level
processed-trial
Analysis outputs
prior-block psychometric fits
Data scope
one OpenAlyx public session
Canonical axis
right contrast positive, left contrast negative

Reproduce locally

Copy this recipe to rebuild the slice's artifacts on your machine at the exact commit deployed here.

Reproduce locally

Run the slice pipeline at this commit

git clone https://github.com/aeronjl/behavtaskatlas.git
cd behavtaskatlas
git checkout ec14ee5
uv sync --extra ibl
uv run behavtaskatlas ibl-visual-decision-download
uv run behavtaskatlas ibl-visual-decision-harmonize
uv run behavtaskatlas ibl-visual-decision-analyze
uv run behavtaskatlas ibl-visual-decision-report

Extras group ibl pulls in the analysis dependencies for this family. Recipe pinned to commit ec14ee5 so a teammate runs the same code path the deploy used.

Artifacts

Slice has a generated local report; the analysis pipeline ran cleanly and produced trial-level output, plots, and provenance.