drop enercast submodule; drop Rcpp requirement; fix doc, complete code, fix fix fix
[epclust.git] / epclust / R / computeSynchrones.R
CommitLineData
40f12a2f
BA
1#' computeSynchrones
2#'
3#' Compute the synchrones curves (sum of clusters elements) from a matrix of medoids,
4#' using euclidian distance.
5#'
6#' @param medoids matrix of medoids in columns (curves of same length as the series)
7#' @param getSeries Function to retrieve series (argument: 'indices', integer vector)
8#' @param nb_curves How many series? (this is known, at this stage)
9#' @inheritParams claws
10#'
11#' @return A matrix of K synchrones in columns (same length as the series)
12#'
13#' @export
282342ba 14computeSynchrones <- function(medoids, getSeries, nb_curves,
40f12a2f
BA
15 nb_series_per_chunk, ncores_clust=1,verbose=FALSE,parll=TRUE)
16{
17 # Synchrones computation is embarassingly parallel: compute it by chunks of series
282342ba 18 computeSynchronesChunk <- function(indices)
40f12a2f
BA
19 {
20 if (parll)
21 {
22 require("bigmemory", quietly=TRUE)
23 requireNamespace("synchronicity", quietly=TRUE)
24 require("epclust", quietly=TRUE)
25 # The big.matrix objects need to be attached to be usable on the workers
26 synchrones <- bigmemory::attach.big.matrix(synchrones_desc)
27 medoids <- bigmemory::attach.big.matrix(medoids_desc)
28 m <- synchronicity::attach.mutex(m_desc)
29 }
30
31 # Obtain a chunk of reference series
282342ba
BA
32 series_chunk <- getSeries(indices)
33 nb_series_chunk <- ncol(series_chunk)
40f12a2f
BA
34
35 # Get medoids indices for this chunk of series
282342ba 36 mi <- assignMedoids(series_chunk, medoids[,])
40f12a2f
BA
37
38 # Update synchrones using mi above, grouping it by values of mi (in 1...K)
39 # to avoid too many lock/unlock
40 for (i in seq_len(K))
41 {
42 # lock / unlock required because several writes at the same time
43 if (parll)
44 synchronicity::lock(m)
282342ba 45 synchrones[,i] <- synchrones[,i] + rowSums(as.matrix(series_chunk[,mi==i]))
40f12a2f
BA
46 if (parll)
47 synchronicity::unlock(m)
48 }
49 NULL
50 }
51
282342ba
BA
52 K <- ncol(medoids)
53 L <- nrow(medoids)
40f12a2f 54 # Use bigmemory (shared==TRUE by default) + synchronicity to fill synchrones in //
282342ba 55 synchrones <- bigmemory::big.matrix(nrow=L, ncol=K, type="double", init=0.)
40f12a2f 56 # NOTE: synchronicity is only for Linux & MacOS; on Windows: run sequentially
282342ba 57 parll <- (parll && requireNamespace("synchronicity",quietly=TRUE)
40f12a2f
BA
58 && Sys.info()['sysname'] != "Windows")
59 if (parll)
60 {
61 m <- synchronicity::boost.mutex() #for lock/unlock, see computeSynchronesChunk
62 # mutex and big.matrix objects cannot be passed directly:
63 # they will be accessed from their description
64 m_desc <- synchronicity::describe(m)
282342ba 65 synchrones_desc <- bigmemory::describe(synchrones)
40f12a2f
BA
66 medoids <- bigmemory::as.big.matrix(medoids)
67 medoids_desc <- bigmemory::describe(medoids)
282342ba
BA
68 # outfile=="" to see stderr/stdout on terminal
69 cl <- parallel::makeCluster(ncores_clust, outfile="")
40f12a2f 70 parallel::clusterExport(cl, envir=environment(),
3c5a4b08 71 varlist=c("synchrones_desc","m_desc","medoids_desc","getSeries"))
40f12a2f
BA
72 }
73
74 if (verbose)
75 cat(paste("--- Compute ",K," synchrones with ",nb_curves," series\n", sep=""))
76
3c5a4b08 77 # Balance tasks by splitting 1:nb_curves into groups of size <= nb_series_per_chunk
282342ba 78 indices_workers <- .splitIndices(seq_len(nb_curves), nb_series_per_chunk)
40f12a2f
BA
79 ignored <-
80 if (parll)
81 parallel::parLapply(cl, indices_workers, computeSynchronesChunk)
82 else
83 lapply(indices_workers, computeSynchronesChunk)
84
85 if (parll)
86 parallel::stopCluster(cl)
87
88 return (synchrones[,])
89}