X-Git-Url: https://git.auder.net/?a=blobdiff_plain;f=epclust%2FR%2Fclustering.R;h=493f90f31c1c20d0cfb591d4acc043ff32505a8a;hb=8702eb86906bd6d59e07bb887e690a20f29be63f;hp=42e894c805eab138e80b038eaa836d78cfa74103;hpb=74f571a72fd63ae92466d944a9ab4a111d177121;p=epclust.git

diff --git a/epclust/R/clustering.R b/epclust/R/clustering.R
index 42e894c..493f90f 100644
--- a/epclust/R/clustering.R
+++ b/epclust/R/clustering.R
@@ -1,58 +1,73 @@
-# Cluster one full task (nb_curves / ntasks series)
-clusteringTask = function(K1, K2, WER, nb_series_per_chunk, indices_tasks, ncores_clust)
+# Cluster one full task (nb_curves / ntasks series); only step 1
+clusteringTask = function(indices, getCoefs, K1, nb_series_per_chunk, ncores)
 {
-	cl_clust = parallel::makeCluster(ncores_clust)
-	#parallel::clusterExport(cl=cl_clust, varlist=c("fonctions_du_package"), envir=environment())
-	indices_clust = indices_task[[i]]
+	cl = parallel::makeCluster(ncores)
+	parallel::clusterExport(cl, varlist=c("getCoefs","K1"), envir=environment())
 	repeat
 	{
-		nb_workers = max( 1, round( length(indices_clust) / nb_series_per_chunk ) )
-		indices_workers = list()
-		for (i in 1:nb_workers)
+		nb_workers = max( 1, floor( length(indices) / nb_series_per_chunk ) )
+		indices_workers = lapply( seq_len(nb_workers), function(i)
+			indices[(nb_series_per_chunk*(i-1)+1):(nb_series_per_chunk*i)] )
+		# Spread the remaining load among the workers
+		rem = length(indices) %% nb_series_per_chunk
+		while (rem > 0)
 		{
-			upper_bound = ifelse( i<nb_workers,
-				min(nb_series_per_chunk*i,length(indices_clust)), length(indices_clust) )
-			indices_workers[[i]] = indices_clust[(nb_series_per_chunk*(i-1)+1):upper_bound]
+			index = rem%%nb_workers + 1
+			indices_workers[[index]] = c(indices_workers[[index]], tail(indices,rem))
+			rem = rem - 1
 		}
-		indices_clust = parallel::parLapply(cl, indices_workers, clusterChunk, K1, K2*(WER=="mix"))
-		# TODO: soft condition between K2 and K1, before applying final WER step
-		if ((WER=="end" && length(indices_clust)==K1) || (WER=="mix" && length(indices_clust)==K2))
+		indices = unlist( parallel::parLapply( cl, indices_workers, function(inds) {
+			require("epclust", quietly=TRUE)
+			inds[ computeClusters1(getCoefs(inds), K1) ]
+		} ) )
+		if (length(indices) == K1)
 			break
 	}
-	parallel::stopCluster(cl_clust)
-	unlist(indices_clust)
-}
-
-# Cluster a chunk of series inside one task (~max nb_series_per_chunk)
-clusterChunk = function(indices, K1, K2)
-{
-	coeffs = getCoeffs(indices)
-	cl = computeClusters(as.matrix(coeffs[,2:ncol(coeffs)]), K1, diss=FALSE)
-	if (K2 > 0)
-	{
-		curves = computeSynchrones(cl)
-		dists = computeWerDists(curves)
-		cl = computeClusters(dists, K2, diss=TRUE)
-	}
-	indices[cl]
+	parallel::stopCluster(cl)
+	indices #medoids
 }
 
 # Apply the clustering algorithm (PAM) on a coeffs or distances matrix
-computeClusters = function(md, K, diss)
+computeClusters1 = function(coefs, K1)
+	cluster::pam(coefs, K1, diss=FALSE)$id.med
+
+# Cluster a chunk of series inside one task (~max nb_series_per_chunk)
+computeClusters2 = function(medoids, K2, getRefSeries, nb_series_per_chunk)
 {
-	if (!require(cluster, quietly=TRUE))
-		stop("Unable to load cluster library")
-	cluster::pam(md, K, diss=diss)$id.med
+	synchrones = computeSynchrones(medoids, getRefSeries, nb_series_per_chunk)
+	medoids[ cluster::pam(computeWerDists(synchrones), K2, diss=TRUE)$medoids , ]
 }
 
 # Compute the synchrones curves (sum of clusters elements) from a clustering result
-computeSynchrones = function(indices)
+computeSynchrones = function(medoids, getRefSeries, nb_series_per_chunk)
 {
-	colSums( getData(indices) )
+	K = nrow(medoids)
+	synchrones = matrix(0, nrow=K, ncol=ncol(medoids))
+	counts = rep(0,K)
+	index = 1
+	repeat
+	{
+		range = (index-1) + seq_len(nb_series_per_chunk)
+		ref_series = getRefSeries(range)
+		if (is.null(ref_series))
+			break
+		#get medoids indices for this chunk of series
+		for (i in seq_len(nrow(ref_series)))
+		{
+			j = which.min( rowSums( sweep(medoids, 2, ref_series[i,], '-')^2 ) )
+			synchrones[j,] = synchrones[j,] + ref_series[i,]
+			counts[j] = counts[j] + 1
+		}
+		index = index + nb_series_per_chunk
+	}
+	#NOTE: odds for some clusters to be empty? (when series already come from stage 2)
+	#      ...maybe; but let's hope resulting K1' be still quite bigger than K2
+	synchrones = sweep(synchrones, 1, counts, '/')
+	synchrones[ sapply(seq_len(K), function(i) all(!is.nan(synchrones[i,]))) , ]
 }
 
-# Compute the WER distance between the synchrones curves
-computeWerDist = function(curves)
+# Compute the WER distance between the synchrones curves (in rows)
+computeWerDists = function(curves)
 {
 	if (!require("Rwave", quietly=TRUE))
 		stop("Unable to load Rwave library")
@@ -89,7 +104,7 @@ computeWerDist = function(curves)
 	{
 		for (j in (i+1):n)
 		{
-			#TODO: later, compute CWT here (because not enough storage space for 32M series)
+			#TODO: later, compute CWT here (because not enough storage space for 200k series)
 			#      'circular=TRUE' is wrong, should just take values on the sides; to rewrite in C
 			num <- filter(Mod(Xcwt4[[i]] * Conj(Xcwt4[[j]])), fcoefs, circular=TRUE)
 			WX <- filter(Mod(Xcwt4[[i]] * Conj(Xcwt4[[i]])), fcoefs, circular=TRUE)