'update'

[epclust.git] / epclust / R / computeWerDists.R

diff --git a/epclust/R/computeWerDists.R b/epclust/R/computeWerDists.R
index aae1cc1..a813b8f 100644 (file)
--- a/epclust/R/computeWerDists.R
+++ b/epclust/R/computeWerDists.R
@@ -3,21 +3,24 @@
 #' Compute the WER distances between the synchrones curves (in columns), which are
 #' returned (e.g.) by \code{computeSynchrones()}
 #'
-#' @param synchrones A big.matrix of synchrones, in columns. The series have same
-#'   length as the series in the initial dataset
+#' @param indices Range of series indices to cluster
 #' @inheritParams claws
+#' @inheritParams computeSynchrones
 #'
-#' @return A distances matrix of size K1 x K1
+#' @return A distances matrix of size K x K where K == length(indices)
 #'
 #' @export
-computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
-       verbose=FALSE,parll=TRUE)
+computeWerDists = function(indices, getSeries, nb_series_per_chunk, nvoice, nbytes, endian,
+       ncores_clust=1, verbose=FALSE, parll=TRUE)
 {
-       n <- ncol(synchrones)
-       L <- nrow(synchrones)
+       n <- length(indices)
+       L <- length(getSeries(1)) #TODO: not very nice way to get L
        noctave = ceiling(log2(L)) #min power of 2 to cover serie range
+       # Since a CWT contains noctave*nvoice complex series, we deduce the number of CWT to
+       # retrieve/put in one chunk.
+       nb_cwt_per_chunk = max(1, floor(nb_series_per_chunk / (nvoice*noctave*2)))
 
-       # Initialize result as a square big.matrix of size 'number of synchrones'
+       # Initialize result as a square big.matrix of size 'number of medoids'
        Xwer_dist <- bigmemory::big.matrix(nrow=n, ncol=n, type="double")
 
        # Generate n(n-1)/2 pairs for WER distances computations
@@ -30,7 +33,7 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
        }
 
        cwt_file = ".cwt.bin"
-       # Compute the synchrones[,indices] CWT, and store the results in the binary file
+       # Compute the getSeries(indices) CWT, and store the results in the binary file
        computeSaveCWT = function(indices)
        {
                if (parll)
@@ -38,27 +41,25 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
                        require("bigmemory", quietly=TRUE)
                        require("Rwave", quietly=TRUE)
                        require("epclust", quietly=TRUE)
-                       synchrones <- bigmemory::attach.big.matrix(synchrones_desc)
                }
 
                # Obtain CWT as big vectors of real part + imaginary part (concatenate)
                ts_cwt <- sapply(indices, function(i) {
-                       ts <- scale(ts(synchrones[,i]), center=TRUE, scale=FALSE)
+                       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)) )
                })
 
                # Serialization
-               binarize(ts_cwt, cwt_file, 1, ",", nbytes, endian)
+               binarize(ts_cwt, cwt_file, nb_cwt_per_chunk, ",", nbytes, endian)
        }
 
        if (parll)
        {
                cl = parallel::makeCluster(ncores_clust)
-               synchrones_desc <- bigmemory::describe(synchrones)
                Xwer_dist_desc <- bigmemory::describe(Xwer_dist)
-               parallel::clusterExport(cl, varlist=c("parll","synchrones_desc","Xwer_dist_desc",
-                       "noctave","nvoice","verbose","getCWT"), envir=environment())
+               parallel::clusterExport(cl, varlist=c("parll","nb_cwt_per_chunk","L",
+                       "Xwer_dist_desc","noctave","nvoice","getCWT"), envir=environment())
        }
 
        if (verbose)
@@ -71,7 +72,7 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
                        lapply(1:n, computeSaveCWT)
 
        # Function to retrieve a synchrone CWT from (binary) file
-       getSynchroneCWT = function(index, L)
+       getCWT = function(index, L)
        {
                flat_cwt <- getDataInFile(index, cwt_file, nbytes, endian)
                cwt_length = length(flat_cwt) / 2
@@ -84,8 +85,6 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
 
 
 #TODO: better repartition here, 
-       #better code in .splitIndices :: never exceed nb_per_chunk; arg: min_per_chunk (default: 1)
-###TODO: reintroduire nb_items_clust ======> l'autre est typiquement + grand !!! (pas de relation !)
 
 
 
@@ -97,7 +96,6 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
                        # parallel workers start with an empty environment
                        require("bigmemory", quietly=TRUE)
                        require("epclust", quietly=TRUE)
-                       synchrones <- bigmemory::attach.big.matrix(synchrones_desc)
                        Xwer_dist <- bigmemory::attach.big.matrix(Xwer_dist_desc)
                }
 
@@ -106,9 +104,8 @@ computeWerDists = function(synchrones, nvoice, nbytes,endian,ncores_clust=1,
                        cat(paste("   Distances (",i,",",j,"), (",i,",",j+1,") ...\n", sep=""))
 
                # Compute CWT of columns i and j in synchrones
-               L = nrow(synchrones)
-               cwt_i <- getSynchroneCWT(i, L)
-               cwt_j <- getSynchroneCWT(j, L)
+               cwt_i <- getCWT(i, L)
+               cwt_j <- getCWT(j, L)
 
                # Compute the ratio of integrals formula 5.6 for WER^2
                # in https://arxiv.org/abs/1101.4744v2 §5.3