Get distance between samples based on bias corrected deviations
getSampleDistance(object, threshold = 1.5, initial_dims = 50, distance_function = dist)
| object | deviations result |
|---|---|
| threshold | threshold for variability |
| initial_dims | initial dimentions for preliminary dimensionality reduction via pca |
| distance_function | distance function to use |
dist object for distance between samples
This function will compute the distance between samples based on the normalized deviations. It will first remove correlated motifs / peak sets. Then the dimensionality will be further reduced via PCA if the number of dimensions exceeds initial_dims. Then the supplied distance_function will be used.
# NOT RUN { # Load very small example results from computeDeviations data(mini_dev, package = "chromVAR") sample_dist <- getSampleDistance(mini_dev, threshold = 0.8) # setting very low variabilitiy threshold because this is mini data set # threshold should generally be above 1 # Use plotVariability to get a sense of an appropriate threshold # As this is mini data set, results not meaningful! # }