Quickstart

using UnfoldBIDS
using Unfold
using LazyArtifacts
using Main: @artifact_str # this is a workaround for Artifacts used in docs; locally you would `using LazyArtifacts`

Loading data

To load use UnfoldBIDS to find the paths to all subject specific data you can uye the bidsLayout function:

sample_data_path = artifact"sample_BIDS"
layout_df = bids_layout(sample_data_path, derivatives=false)

3×6 DataFrame

Row	subject	ses	task	run	file	events
	Any	Any	Any	Any	Any	Any
1	001	missing	P3	missing	/home/runner/.julia/artifacts/6fa05e5d4476b16de1d27962bf672c119fa05c87/sample_ds/sub-001/eeg/sub-001_task-P3_eeg.set	sub-001_task-P3_events.tsv
2	002	missing	P3	missing	/home/runner/.julia/artifacts/6fa05e5d4476b16de1d27962bf672c119fa05c87/sample_ds/sub-002/eeg/sub-002_task-P3_eeg.set	sub-002_task-P3_events.tsv
3	003	missing	P3	missing	/home/runner/.julia/artifacts/6fa05e5d4476b16de1d27962bf672c119fa05c87/sample_ds/sub-003/eeg/sub-003_task-P3_eeg.set	sub-003_task-P3_events.tsv

This will give you a DataFrame containing the paths too the eeg files of all subjects plus their accompanying event files

Note

Since we set the derivative keyword to false here UnfoldBIDS will only look for the raw EEG files. However, by default UnfoldBIDS assumes you have preprocessed data in a derivatives folder and try to look for those.

Subsequently, you can load the data of all subjects into memory

data_df = load_bids_eeg_data(layout_df)

3×6 DataFrame

Row	subject	ses	task	run	raw	events
	SubStrin…	Missing	SubStrin…	Missing	Py	DataFrame
1	001	missing	P3	missing	<RawEEGLAB \| sub-001_task-P3_eeg.fdt, 33 x 478208 (467.0 s), ~41 kB, data not loaded>	402×6 DataFrame
2	002	missing	P3	missing	<RawEEGLAB \| sub-002_task-P3_eeg.fdt, 33 x 414720 (405.0 s), ~41 kB, data not loaded>	402×6 DataFrame
3	003	missing	P3	missing	<RawEEGLAB \| sub-003_task-P3_eeg.fdt, 33 x 386048 (377.0 s), ~41 kB, data not loaded>	401×6 DataFrame

Note

At this point in time, the data is not yet actually loaded into memory, but uses MNE's lazy loading functionality.

As you can see, UnfoldBIDS trys to load events directly into the DataFrame, however if you are missing the event tsv files you will get a warning and no events are loaded. If that happens you have to manually load these events. The following function might help you with this. (The resulting dataframe still needs to be added to data_df!)

events_df = load_events(layout_df)

Run unfold type models

basisfunction = firbasis(τ=(-0.2,.8),sfreq=1024)
f  = @formula 0~1
bfDict = ["stimulus"=>(f,basisfunction)]
UnfoldBIDS.rename_to_latency(data_df, :sample) # Unfold expects a :latency collumn in your events; if your event latency is named differently you can use this function as remedy

resultsAll = run_unfold(data_df, bfDict; eventcolumn="trial_type");


33.3%┣██████████████▍                            ┫ 1/3 [00:00<Inf:Inf, InfGs/it]

Progress:   6%|██▌                                      |  ETA: 0:00:12
Progress: 100%|█████████████████████████████████████████| Time: 0:00:01

66.7%┣███████████████████████████████▍               ┫ 2/3 [00:12<00:12, 12s/it]

Progress:  58%|███████████████████████▋                 |  ETA: 0:00:00
Progress: 100%|█████████████████████████████████████████| Time: 0:00:00

100.0%┣███████████████████████████████████████████████┫ 3/3 [00:12<00:00, 6s/it]

Progress:  67%|███████████████████████████▍             |  ETA: 0:00:00
Progress: 100%|█████████████████████████████████████████| Time: 0:00:00

This page was generated using Literate.jl.