Skip to content

Files

Latest commit

e4c4d60 · Jan 30, 2025

History

History
46 lines (29 loc) · 2.11 KB

NEWS.md

File metadata and controls

46 lines (29 loc) · 2.11 KB

momimi 0.0.1

  • Adding a NEWS.md file to track changes.

momimi 0.0.2

  • Adding the 3-parameter version of TMM (free parameters are the relative motor execution threshold, the peak time, and the curvature).

momimi 0.0.3

  • Adding the possibility to include within-trial (diffusive) noise.
  • Fixing the 3-parameter version of the TMM (TMM3).

momimi 0.0.4

  • Fixing the plotting method of momimi_sim.
  • It is now possible to estimate the amount of between-trial variability (noise) in the TMM3 (which thus now has 4 free parameters).

momimi 0.0.5

  • Fixing some erroneous labelling in the plotting utilities.
  • Adding the quantiles option in plotting methods for fitted objects (qq-plot).
  • Adding the possibility to fit models with function-level noise or diffusive noise (in addition to only parameter-level noise). These two options are slower because it requires numerically finding the RT and MT.
  • The 3-parameter version of the TMM (TMM3) now really has only 3 parameters, whereas the 4-parameter version now has the motor execution threshold, the peak time, the curvature, and the amount of between-trial variability as free parameters.
  • It is now possible to fit the models with "brute force" (i.e., defining a discrete grid and looking for the minimum).

momimi 0.0.6

  • Fixing some erroneous labelling in the plotting utilities.
  • Now returning the full error surface in grid_search fitting method.
  • Removing constraints during fitting (i.e., exec_threshold was previously expected between 0.25 and 4).

momimi 0.0.7

  • Removing the parallel inhibition model (non-identifiable).
  • Simplifying the threshold modulation model by dropping the amplitude parameter (fixed to 1).

momimi 0.0.8

  • Adding the possibility to specify the number of available cores in fitting().

momimi 0.0.9

  • Fixing some typos.
  • Fixing an error related to the presence of a balance variable (defined in a previous version of the model).
  • Removed the 3-parameter version of the TMM: now the number of free parameters can simply be adjusted by modifying the lower_bounds and upper_bounds arguments when using the fitting() function.