EEGBrowser Workflows#

EEGPrep includes a scrolling browser for inspecting continuous, epoched, component, spectral, and overlay data. The browser is available from the Plot menu, rejection workflows, Python APIs, and eegprep-console.

Opening The Browser#

From the GUI, load or select a dataset and use the Plot menu entries for channel data, component activity, or visual rejection. The rejection dialogs can also open browser-backed review windows before updating marks or removing data.

With the tutorial data:

Load sample_data/eeglab_data.set.

Choose Plot > Channel data (scroll).

Use the browser to review channels, events, scaling, and visible time ranges before making rejection decisions.

Run LASTCOM in eegprep-console to confirm which browser action was recorded.

From Python or eegprep-console:

from eegprep import eegplot, pop_eegplot

window = eegplot(EEG)
EEG, com = pop_eegplot(EEG, return_com=True)

Use show=False when testing or scripting normalization without opening Qt:

model = eegplot(EEG, winlength=5, dispchans=16, show=False)

Browser Modes#

Workflow	GUI entry	Programmatic entry
Channel inspection	`Plot > Channel data (scroll)`	`eegplot(EEG)` or `pop_eegplot(EEG)`
Component inspection	`Plot > Component activations (scroll)`	`pop_eegplot(EEG, icacomp=0, reject=0)`
Epoched data rejection	`Tools > Reject data epochs > Reject by inspection`	`pop_eegplot(EEG, reject=1)`
ICA rejection marks	`Tools > Reject data using ICA > Reject by inspection`	`pop_eegplot(EEG, icacomp=0, reject=1)`

Marks And Rejection#

Dragging across the browser marks a time range. On continuous data, accepted marks can be stored in EEG.reject.rejmanualwinrej for later review or converted through eeg_eegrej to remove samples. On epoched data, accepted marks update EEG.reject.rejmanual and EEG.reject.rejmanualE or remove marked epochs through pop_rejepoch.

Component mode uses ICA activations and writes ICA-prefixed rejection fields where applicable. Event overlays use EEGLAB one-based event latencies; Python array indexing remains zero-based internally.

Accepting Browser Results#

In eegprep-console, browser accept callbacks keep GUI state, EEG, ALLEEG, and history synchronized. If a browser action creates a new dataset instead of mutating the current one, the session stores that dataset through the same dataset helpers used by menu actions.

In normal Python, call the lower-level rejection helpers explicitly after reviewing the browser model or marks.

Performance#

The browser draws only the visible time range and applies min/max decimation to large traces, preserving endpoints and extrema while keeping UI redraw work bounded by the displayed pixel width.