Dataset

Processed Roitman-Shadlen random-dot motion reaction-time data from PyDDM

Processed CSV copy of the Roitman and Shadlen random-dot motion reaction-time task data distributed with the PyDDM documentation. The table contains monkey identifier, reaction time, unsigned coherence, correctness, and target choice for two macaques.

Access notes

The MVP downloads the processed CSV from a pinned PyDDM repository commit into ignored local raw-data storage. The original dataset is linked by PyDDM and eLife to the Shadlen Lab RoitmanDataCode resource.

Data formats

  • CSV
  • derived canonical trial tables to be generated

Expected trial-table mapping

Canonical fields the harmonizer expects this dataset to populate.

Canonical field Source mapping
choice `trgchoice` mapped as target 1/right and target 2/left.
correct `correct` mapped to boolean correctness.
evidence_strength Absolute percent motion coherence.
response_time `rt` in seconds.
session_id Processed dataset identifier because session labels are not retained in the convenience CSV.
stimulus_side Target 1 mapped to right and target 2 mapped to left as a canonical convention.
stimulus_value Signed motion coherence, target 1 positive and target 2 negative.
subject_id Monkey identifier formatted from the `monkey` column.
task_variables Raw monkey id, unsigned coherence, signed coherence fraction, target choice, and correctness code.
trial_index Row index within the processed CSV after optional filtering.

Caveats

  • The convenience CSV does not preserve original session labels.
  • The convenience CSV does not preserve stable screen-side mapping for target 1 and target 2.
  • Original dataset redistribution terms should be confirmed before committing derived tables.

Papers

Findings

Curves attached to this dataset on shared signed-evidence axes.

See all findings overlaid on shared axes.