Update history

09/04/2026: Added an introduction to TRF.

07/12/2026: [Version 0.2.9] Added make_predictors_from_tables() to trf module to allow generation of impulse-like predictors using tables or spreadsheets. See the Predictors from tables tab for usage.

07/10/2026: [Version 0.2.8] Added make_pitch_predictors() to trf module to extract pitch contours and generate pitch predictors. This new functionality is implemented using praat-parselmouth so you would need to install it or update your neuraspeech environment using the updated environment.yml. See the Pitch predictors tab for usage.

06/29/2026: [Version 0.2.7] Added make_predictors_from_textgrids() to trf module to allow predictor generation using TextGrid annotations of the audio stimuli.

06/11/2026: [Version 0.2.6] Expanded prepare_eeg() functionality of trf module so that it can now handles EEG epochs containing more than one stimulus. The plotting functions in the prp module now also support plotting standard errors around the mean curve.

The TRF documentation has also been updated. It now covers the complete workflow for running neural tracking of acoustics as well as basic result analyses.

05/29/2026: Added documentation for running TRF models and submitting Slurm jobs.

05/16/2026: [Version 0.2.5] Updated the plot_schematic() function to the trf module for plotting EEG, stimulus waveform, and predictors.

05/13/2026: [Version 0.2.4] Updated the prepare_eeg function to allow concatenation of EEG epochs and predictors.

05/04/2026: Started adding documentation and tutorials for the trf module.

05/03/2026: [Version 0.2.3] A few updates to the boosting script for estimating TRFs:

  • From now on the user will specify the models to be estimated by listing the name of each model and its corresponding predictors. This replaces the old way of specifying a base model and then setting the estimation model ('base_only' or 'unique_contribution').
  • The user can now use include or exclude to subset the data files to process.
  • Added the skip_estimated_models. If True, TRF estimation won't be run again for a given model if the results file has already existed.

Also updated make_gammatone_predictors to allow using a custom predictor name other than 'gammatone'.

04/22/2026: [Version 0.2.2] Core functions of the trf module now accept keyword arguments in addition to a configuration file (detailed usage notes will be uploaded). Also updated the example job script.

04/17/2026: [Version 0.2.1] Added a new make_predictors_from_textgrids function for generating predictors from TextGrid annotations of the stimuli, and fixed a few things:

  • Updated environment.yaml to include textgrids package installation
  • Fixed bugs in the boosting script for TRF estimation and the example job script

04/07/2026: [Version 0.2.0] Now including a trf module for running TRF analysis. A quick guide for using it is available. More detailed documentation will be provided in the future.

03/24/2026: Added a get_values() function for more convinently getting PRP amplitude and F-statistic values at specific time points/intervals and at specific electrodes. See Extracting PRP amplitudes and Extracting F-statistics in the PRP basics and F-statistic sections.

01/15/2026: Added a combine() utility to concatenate multiple datasets (see PRP basics); fixed a bug with recode().