Clustering and supplementary analyses
dg_walktrap_ER.R Takes trimmed data from RedCap and computes Walktrap clustering analysis to determine clusters in the data. Optimizes the modularity result by selecting the number of random steps that yields the highest modularity. Lastly, plots and saves results. Note: For now, the script works for 2 clusters, but contact David to make it work for any number of clusters.
Based on Mackenzie's Walktrap_ExampleCode_MEM.R
Needs dg_determine_groups.R in the same directory.
dg_itemlevel_data_converter_ER.R Given an Excel with labels document downloaded from RedCap, it processes the data, and recodes all the items of several subscores by finding the least common multiple (LCM). The script can handle any range of item-level data and number of subscores. Lastly, it saves the rescaled data in a .rds file.
dg_trim_clinicaldata.R The script cleans and organizes the data collected with 3 surveys used in the BrainMAP study.
dg_determine_age.R The script runs dg_trim_clinicaldata.R and reads the frequently updated "Demographics Form.csv" BrainMAP document to organize the surveys' timestamps, dates of birth, and calculate age per subject.
dg_determine_groups.R The script reads the "Demographics Form.csv" document and provides a list of participants regarded as with ADHD or as TD. The Demographics Form.csv is found in SharePoint: cohenlabteam/Documents/Research Studies/ADHD BrainMAP/Data/
dg_MWUtest_afterWalktrap.R The script compares 2 clusters of BrainMAP participants (either ADHD only or ADHD plus TD) using a Wilcoxon rank sum test, equivalent to Mann-Whitney U test. Lastly, it makes violin plots to optionally check for the data distribution. Used for preliminary results.
Needs dg_determine_groups.R and dg_trim_clinicaldata.R.
¡¡It also needs the converted/rescaled data with dg_itemlevel_data_converter_ER.R!!