X-Git-Url: https://git.auder.net/?a=blobdiff_plain;f=epclust%2FR%2FcomputeWerDists.R;h=f47cd238767cbab18eca0ccec1b8121eec96404f;hb=57f337af19cd6251815bb1ff2d62f4c58e8b6078;hp=8eb755c9920d86f2934d15eb90ec560ba4f596e7;hpb=dc86eb0c992e6e4ab119d48398d040c4cf3a75fd;p=epclust.git

diff --git a/epclust/R/computeWerDists.R b/epclust/R/computeWerDists.R
index 8eb755c..f47cd23 100644
--- a/epclust/R/computeWerDists.R
+++ b/epclust/R/computeWerDists.R
@@ -3,8 +3,8 @@
 #' Compute the WER distances between the series at specified indices, which are
 #' obtaind by \code{getSeries(indices)}
 #'
-#' @param indices Range of series indices to cluster
-#' @param getSeries Function to retrieve series (argument: 'indices', integer vector),
+#' @param indices Indices of the series to consider
+#' @param getSeries Function to retrieve series (argument: 'inds', integer vector),
 #'   as columns of a matrix
 #' @param ncores Number of cores for parallel runs
 #' @inheritParams claws
@@ -12,8 +12,8 @@
 #' @return A distances matrix of size K x K where K == length(indices)
 #'
 #' @export
-computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl, nvoice,
-	nbytes, endian, ncores=3, verbose=FALSE, parll=TRUE)
+computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl=3,
+	nvoice=4, nbytes=4, endian=.Platform$endian, ncores=3, verbose=FALSE)
 {
 	n <- length(indices)
 	L <- length(getSeries(1)) #TODO: not very neat way to get L
@@ -27,10 +27,13 @@ computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl,
 
 	cwt_file <- tempfile(pattern="epclust_cwt.bin_")
 	# Compute the getSeries(indices) CWT, and store the results in the binary file
-	computeSaveCWT <- function(indices)
+	computeSaveCWT <- function(inds)
 	{
+		if (verbose)
+			cat("   Compute save CWT on ",length(inds)," indices\n", sep="")
+
 		# Obtain CWT as big vectors of real part + imaginary part (concatenate)
-		ts_cwt <- sapply(indices, function(i) {
+		ts_cwt <- sapply(inds, function(i) {
 			ts <- scale(ts(getSeries(i)), center=TRUE, scale=FALSE)
 			ts_cwt <- Rwave::cwt(ts, noctave, nvoice, w0=2*pi, twoD=TRUE, plot=FALSE)
 			c( as.double(Re(ts_cwt)),as.double(Im(ts_cwt)) )
@@ -60,7 +63,7 @@ computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl,
 			Xwer_dist <- bigmemory::attach.big.matrix(Xwer_dist_desc)
 		}
 
-		if (verbose && !parll)
+		if (verbose)
 			cat(paste("   Distances from ",i," to ",i+1,"...",n,"\n", sep=""))
 
 		# Get CWT of column i, and run computations for columns j>i
@@ -87,13 +90,14 @@ computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl,
 		cat(paste("--- Precompute and serialize synchrones CWT\n", sep=""))
 
 	# Split indices by packets of length at most nb_cwt_per_chunk
-	indices_cwt <- .splitIndices(seq_len(n), nb_cwt_per_chunk)
+	indices_cwt <- .splitIndices(indices, nb_cwt_per_chunk)
 	# NOTE: next loop could potentially be run in //. Indices would be permuted (by
 	# serialization order), and synchronicity would be required because of concurrent
-	# writes. Probably not worth the effort - but possible to gain some bits of speed.
+	# writes. Probably not worth the effort - but possible.
 	for (inds in indices_cwt)
 		computeSaveCWT(inds)
 
+	parll <- (ncores > 1)
 	if (parll)
 	{
 		# outfile=="" to see stderr/stdout on terminal
@@ -112,7 +116,7 @@ computeWerDists <- function(indices, getSeries, nb_series_per_chunk, smooth_lvl,
 
 	ignored <-
 		if (parll)
-			parallel::parLapply(cl, seq_len(n-1), computeDistances)
+			parallel::parLapplyLB(cl, seq_len(n-1), computeDistances)
 		else
 			lapply(seq_len(n-1), computeDistances)
 	Xwer_dist[n,n] <- 0.