douwehorsthuis/eeg_to_erp_pipeline_stats_r

General pipeline used for analyzing EEG data where Raw EEG data gets transformed into ERPS and Stats are done in R (Mixed effects models)

This is the Github page of Douwe Horsthuis

• 🌱 I’m working with SQL, Matlab, R, Presentation (Scenario Description Language (SDL) & Presentation Control Language (PCL)) (Github calls this language scilabs), Rmarkdown/markdown (see the readme files) and always trying to learn more about anything I work with
• 🧑‍🔬 Currently working as a data analyst
• 👯 I’m looking to collaborate on socially meaningful project
• 📫 How to reach me: [email protected]
• 😄 Pronouns: he/his/him
• 🤖 check out my website www.douwehorsthuis.com

I am a Data analyst at the Belastingdienst-dienst toeslagen (in the Netherlands). I used to be a lab manager and EEG technician in the CNL at Albert Einstein College of Medicine in the Bronx. Most of the code on this page is related to my work at the The Cogntive Neurophysiology Laboratory and is availible on CognitiveNeuroLab's github page.

I work with different languages and platforms best exemplified by these repos:
Matlab & RMarkdown
R & RMarkdown
Markdown
Presentation (Scenario Description Language (SDL) & Presentation Control Language (PCL)) (Github calls this language scilabs)
And I am familiar with Tableau.
I am currently mainly coding in SAS and SQL.

eeg_to_erp_pipeline_stats_r's Issues

re-referencing order

if re-referencing to a specific channel, that should be done after ICA

finding functions in script 2

add the function folder to path automatic

update to include group quality

add to the C-manual_check script:

quality(s,:)=[str2double(subject_list{s}), EEG.deleteddata_wboundries, length(EEG.del_chan)];

pop_clean_rawdata, not excluding externals from cleaning

The pop_clean_rawdata function is able to exclude externals.
See sccn/clean_rawdata#28

move plot bad channels to manual check script

should be there because that's where ultimatly the final amount of channels always exist

Checking for bridged channels

the new cleaning function looks for correlations, but we need to check if it checks for min or max correlations.

EEG.event problem

figure out why EEG.event created edftypes with condition and without

bridge function in wrong location

need to move it down to when the data has channel locations

Double check how often you re-reference

@anafrancisco mentioned that there are several moments where a avg reference and even a tp8 reference happend. This should not be the case