R/transposeBigData.R

In milescsmith/WGCNA: Weighted Correlation Network Analysis

transposeBigData <- function(x, blocksize = 20000) {
  isdataframe <- is.data.frame(x)
  ismatrix <- is.matrix(x)
  if (!(isdataframe | ismatrix)) {
    stop("Input is neither a data frame nor a matrix")
  }
  if (blocksize < 2) {
    stop("This blocksize makes no sense. It should be a positive integer>1.")
  }
  nrow1 <- nrow(x)
  ncol1 <- ncol(x)
  xTranspose <- matrix(NA, nrow = ncol1, ncol = nrow1)
  if (nrow1 <= ncol1) {
    no.blocks <- as.integer(ncol1 / blocksize)
    if (no.blocks >= 1) {
      for (i in 1:no.blocks) {
        blockIndex <- (i - 1) * blocksize + 1:blocksize
        xTranspose[blockIndex, ] <- t(x[, blockIndex])
      }
    }
    if (ncol1 - no.blocks * blocksize == 1) {
      xTranspose[ncol1, ] <- t(x[, ncol1])
    }
    if (ncol1 - no.blocks * blocksize > 1) {
      finalindex <- (no.blocks * blocksize + 1):ncol1
      xTranspose[finalindex, ] <- t(x[, finalindex])
    }
  }
  if (nrow1 > ncol1) {
    no.blocks <- as.integer(nrow1 / blocksize)
    if (no.blocks >= 1) {
      for (i in 1:no.blocks) {
        blockIndex <- (i - 1) * blocksize + 1:blocksize
        xTranspose[, blockIndex] <- t(x[blockIndex, ])
      }
    }
    if (nrow1 - no.blocks * blocksize == 1) {
      xTranspose[, nrow1] <- t(x[nrow1, ])
    }
    if (nrow1 - no.blocks * blocksize > 1) {
      finalindex <- (no.blocks * blocksize + 1):nrow1
      xTranspose[, finalindex] <- t(x[finalindex, ])
    }
  }
  if (isdataframe) {
    xTranspose <- data.frame(xTranspose)
    dimnames(xTranspose)[[1]] <- dimnames(x)[[2]]
    dimnames(xTranspose)[[2]] <- dimnames(x)[[1]]
  }
  if (ismatrix) {
    dimnames(xTranspose)[[1]] <- dimnames(x)[[2]]
    dimnames(xTranspose)[[2]] <- dimnames(x)[[1]]
  }
  xTranspose
}