Matrix convert to binary

This function takes a weighted adjacency matrix and separates the components of the matrix into weights and presence/absence matrices making use of a threshold.

network_make_binary(adj_mat, threshold, print_message = TRUE)

Arguments

adj_mat	adjacency matrix to be converted.
threshold	we recommended to use network_rescale before using this function. Re-scaling will transform all values into a range [0,100]. The threshold can either be a numeric balue in the interval [0,100] or a character string specifying the following methods for automatically determining the threshold based on the input data: threshold = "median": compute the median over the entire input adj_mat and use this median value as threshold for defining all edge weights of a genes equal or below the median value as 0 and all values above the median value as 1.
print_message	a logical value indicating whether or not a threshold message shall be printed.

Author

Sergio Vasquez and Hajk-Georg Drost

Examples

 # path to PPCOR output file
ppcor_output <- system.file('beeline_examples/PPCOR/outFile.txt', package = 'edgynode')
 # import PPCOR specific output
ppcor_parsed <- ppcor(ppcor_output)
 # rescale the matrix
ppcor_rescaled <- network_rescale(ppcor_parsed)
#> It seems like your input matrix contains values of correlation coefficients range(-1,1). Please be aware that for negative values the absolute value will be taken before rescaling. We transformed all negative values to their absolute values.
 # make the binary matrix
ppcor_binary <- network_make_binary(ppcor_rescaled, 70)
#> network_make_binary() applies [70] as cut-off threshold to transform the input weighted adjacency matrix into a binary adjacency matrix.
 # to see the differences you may use:
 plot_network_dendrogram(ppcor_rescaled)
 plot_network_dendrogram(ppcor_binary)