Data Products document,
which contains descriptions of both the raw
data (this is primarily meant for internal documentation purposes) and the
derived data products (this is meant for public consumption, to aid users
of the derived data products in performing their analyses).
New version of derived data products. The new data process introduces
substantial changes, both from the user's standpoint and under the hood.
From the user perspective, the new process includes additional experiments
and raw channels of fNIRS that were previously excluded. It also integrates
a comprehensive report on the EEG and fNIRS filtering process, combines the
entire process report into a single file, offers CSV files with filtered
and synchronized signals, and removes NaN values in signal columns. Naming
has been improved with nirs replaced by fnirs, and the experiment_id
column has been eliminated from CSV files. Internally, the signal
processing methods for EEG and fNIRS signals have been updated, including
changes to grouping, interpolation, filtering, and resampling approaches.
Signals are now synchronized and resampled before task grouping, linear
interpolation is used for consistent signal intervals before filtering, the
Butterworth filter order is reduced to 4, the Chebychev type I filter
replaces the lowpass FIR filter for aliasing avoidance, and a fast Fourier
transform-based method is used for resampling to target frequency.
Additionally, some tweaks have been made to the website layout and design.
nirsreplaced byfnirs, and theexperiment_idcolumn has been eliminated from CSV files. Internally, the signal processing methods for EEG and fNIRS signals have been updated, including changes to grouping, interpolation, filtering, and resampling approaches. Signals are now synchronized and resampled before task grouping, linear interpolation is used for consistent signal intervals before filtering, the Butterworth filter order is reduced to 4, the Chebychev type I filter replaces the lowpass FIR filter for aliasing avoidance, and a fast Fourier transform-based method is used for resampling to target frequency.Additionally, some tweaks have been made to the website layout and design.