MIC

Multilevel Integrative Clustering, packaged by Qian Li

Installation:

devtools::install_github("Qian-Li/MIC")

See our manuscript for more details!

Quick Example

Time series simulation:

ts_sim <- MIC_sim(alpha = 0.9, nsub = 3, segs = 10, fs = 100)

Data preparation:

# Spectral estimation on simulated time series.
list_data <- lapply(ts_sim$Data, function(x) MIC_prep(X = x, d = 4, 
  par.spectrum = c(50, 50, 100), par.win = c(3, 1)))

MIC modle fitting:

# Running time approx. 3 mins
output <- MIC(X = list_data, K = 4, NumRun = 5000)

Clustering result:

# Accuracy relative to the true group labels
sum(clust_align(ts_sim$C, output$Cbest, type = 'vec') == ts_sim$C) / 40