If two deconvolution features within a specific cell type are found to be highly correlated, one feature is kept randomly for further analysis.
Usage
removeCorrelatedFeatures(
data,
threshold,
name,
n_seed,
corr_method = "spearman"
)
Arguments
- data
Deconvolution matrix
- threshold
Threshold for defined high correlated features
- name
Cell type name corresponding to the given matrix in 'data'
- n_seed
Seed to ensure reproducibility regarding the choice of the feature.
- corr_method
Correlation type whether "spearman" or "pearson".
