Wsws_timefrq.R models subjective clarity ratings and mean spectral power estimates for each analysed frequency band. Note that rerunning analysis pipelines will overwrite these files, potentially altering model results if new parameters have been specified. The outputs tables generated from the spectral power ( wsws_timefrq.csv) and stimulus reconstruction ( wsws_stimrec.csv) analyses have been provided in the stats subdirectory so that statistical modelling can be performed in R without needing to run any component of the MATLAB pipeline. Note that this analysis is independent of the EEG pipeline and can therefore be performed at any time. The pupillometry data are preprocessed and analysed by running wsws_pupil_ERP.m. This pipeline can be run from the high-level script wsws_run_stimrec.m. The stimulus reconstruction analysis can be performed independently of the spectral analysis. Wsws_prep_EEG_LMM.m collates trial-level spectral power estimates for statistical modelling. Grand-average time-frequency compositions can be computed and plotted with wsws_average_EEG_TF.m.
(Note that these scripts can be run on the preprocessed files provided in the OSF repository, thereby obviating the need to run the preprocessing pipeline first.)
The time-frequency and event-related (de)synchronisation analyses can be run from the scripts contained within the EEGtimefrq subdirectory.įor both kinds of analysis, first extract frequency-domain estimates using wsws_extract_EEG_*.m, before conducting cluster-based permutation analysis using wsws_cluster_EEG_*.m. ICs that are marked for rejection can be altered by updating the contents of wsws_badComps_t*.mat prior to running wsws_ICA_reject_t*.m. You may also manually review the ICA decompositions produced by wsws_ICA.m by invoking the wsws_ICA_plot.m function, specifying the appropriate subject number and file type ('train' or 'test'). Note that this will modify the wsws_bad*.csv files that are invoked when the pipeline is run in the default (automated rejection) mode. You may manually review the raw data for bad channels and epochs by setting manualRej = 1. Note that running this script will mean that bad channels, epochs, and independent components will be rejected according to the information stored in the corresponding wsws_bad*.csv and wsws_badComps_t*.mat files. The entire preprocessing pipeline can be conveniently run from the high-level script wsws_run_preproc.m, which will invoke each subcomponent of the pipeline in sequence (refer to this script to ensure correct ordering of operations if running scripts individually). from raw data), first navigate to the EEGpreproc subdirectory to perform preprocessing. To reproduce the EEG analysis pipeline from scratch (i.e. Optional toolboxes for figure plotting include EEGLAB and export_fig. You will need FieldTrip installed in order to reproduce the EEG-based analyses the mTRF toolbox is required for the stimulus reconstruction analysis only (scripts will warn you if these toolboxes cannot be located).
#Current symbolic math toolbox update download#
You will need to modify this script in order to define local paths to the data and stimulus materials (which you will need to download from the OSF repository), as well as the requisite MATLAB toolboxes. In order to reproduce the analyses reported in the manuscript, first download/clone this repository into a suitable directory and navigate to the def_local.m script. You can access the data files on which these scripts operate from the accompanying OSF repository. This repository contains the scripts and functions required to reproduce the analyses reported in the associated manuscript. Expectations boost the reconstruction of auditory features from electrophysiological responses to noisy speech, a combined EEG-pupillometry study of degraded speech processing.
0 kommentar(er)
