R/sample_FBN.R

In clsdavid/FBNNet2_public: Fundamental Boolean Model and Networks

#' This method is used to converts normalized
#' timeseries data into a list of samples
#' @param normalizedData An output of the method 
#' normalizeTimesereisRawData that normalized timeseries data
#' @param func A function that specified how to split the column names
#' @param splitor A separator.
#'
#'@export
convertIntoSampleTimeSeries <- function(normalizedData, func = function(x) paste(x[1], x[2], x[3], sep = "-"), splitor = as.character("-")) {
    
    mat <- normalizedData
    # Obtain the last part of each column names
    groups <- vapply(strsplit(x = colnames(mat), split = splitor), func, character(1))
    print(groups)
    # Go through each unique column name and extract the corresponding columns
    res <- lapply(unique(groups), function(x) mat[, which(groups == x)])
    names(res) <- unique(groups)
    
    # need to convert column name into numbers str_extract_all(names(df), '[0-9]+')
    res
}

#'This method is used to sort time series order based on columns
#'@param convertedTimeSeries An output of the method 
#'convertIntoSampleTimeSeries that converts normalized timeseries data
#' into a list of samples
#'@param func A function that specified how to split the column names
#'@param splitor Separator.
#'@export
reorderSampleTimeSeries <- function(convertedTimeSeries, func = function(x) x[length(x)], splitor = as.character("-")) {
    
    len <- length(convertedTimeSeries)
    res <- lapply(seq_len(len), function(index) {
        result <- convertedTimeSeries[[index]]
        ocolnames <- colnames(result)
        ocolnames <- gsub("\\..*", "", ocolnames)
        ocolnames <- vapply(strsplit(x = ocolnames, split = splitor), func, character(1))
        ocolnames <- as.integer(stringr::str_extract_all(ocolnames, "[0-9]+"))
        colnames(result) <- ocolnames
        result2 <- result[, order(as.numeric(colnames(result)))]
        result2 <- result2[order(rownames(result2)), ]
        return(result2)
    })
    names(res) <- names(convertedTimeSeries)
    res
}