1 ## File : 03_compute-sum-of-classes_2009.r
2 ## Description : Calculer les synchrones pour chaque groupe obtenu par le
7 MOJARRITA <- Sys.info()[4] == "mojarrita"
10 setwd("~/Documents/projects/2014_EDF-Orsay-Lyon2/codes/")
12 setwd("~/2014_EDF-Orsay-Lyon2/codes/")
16 ## 1. Read auxiliar data files ####
18 identifiants <- read.table("identifs.txt")[ ,1]
19 dates0 <- read.table("datesall.txt")[, 1]
20 dates <- dates0[grep("2009", dates0)]
23 n <- length(identifiants)
27 blocks <- c(rep(6500, 3), 5511)
32 # Fit of the clustering : clfit
33 load('../res/clfitdf200.Rdata') # Loads res that containts
34 # clusterings memberships
35 res <- as.data.frame(res)
38 K <- 200 #nrow(clfit$clusinfo)
41 ## 2. Process the large file ####
44 con <- file("~/tmp/2009_full.txt") # Establish a connection to the file
45 open(con, "r") # Open the connection
47 for(b in seq_along(blocks)){ # Reading loop
49 actual <- readLines(con = con, n = nb )
50 auxmat <- matrix(unlist(strsplit(actual, " ")), ncol = p + 1, byrow = TRUE)
53 datamat <- t(apply(auxmat[, -1], 1, as.numeric)) # the NA introduced by as.numeric
55 rownames(datamat) <- substr(auxmat[, 1], 2, 7)
58 synchros <- lapply(res,
60 aux <- matrix(0, ncol = p, nrow = K)
62 clustk <- which(ll == k)
63 if(length(clustk) > 1) {
64 aux[k, ] <- colSums(datamat[ll == k, ])
66 aux[k, ] <- datamat[ll == k, ]
73 close(con) # close connection to the file
75 # save(synchros, file = "~/tmp/2009synchrosdf200WER")