projects / epclust.git / blame
| Commit | Line | Data |
|---|---|---|
| ad642dc6 BA |
1 | ## File : 03_compute-sum-of-classes_2009.r |
| 2 | ## Description : Calculer les synchrones pour chaque groupe obtenu par le | |
| 3 | ## clustering. | |
| 4 | ||
| 5 | rm(list = ls()) | |
| 6 | ||
| 7 | MOJARRITA <- Sys.info()[4] == "mojarrita" | |
| 8 | ||
| 9 | if(MOJARRITA){ | |
| 10 | setwd("~/Documents/projects/2014_EDF-Orsay-Lyon2/codes/") | |
| 11 | } else { | |
| 12 | setwd("~/2014_EDF-Orsay-Lyon2/codes/") | |
| 13 | } | |
| 14 | ||
| 15 | ||
| 16 | ## 1. Read auxiliar data files #### | |
| 17 | ||
| 18 | identifiants <- read.table("identifs.txt")[ ,1] | |
| 19 | dates0 <- read.table("datesall.txt")[, 1] | |
| 20 | dates <- dates0[grep("2009", dates0)] | |
| 21 | rm(dates0) | |
| 22 | ||
| 23 | n <- length(identifiants) | |
| 24 | p <- length(dates) | |
| 25 | ||
| 26 | if(MOJARRITA) { | |
| 27 | blocks <- c(rep(6500, 3), 5511) | |
| 28 | } else { | |
| 29 | blocks <- 25011 | |
| 30 | } | |
| 31 | ||
| 32 | # Fit of the clustering : clfit | |
| 33 | load('../res/clfitdf200.Rdata') # Loads res that containts | |
| 34 | # clusterings memberships | |
| 35 | res <- as.data.frame(res) | |
| 36 | ||
| 37 | resRDN <- as.data.frame(lapply(res, sample)) | |
| 38 | #save(file = "../res/clfitdf200RDN.Rdata", resRDN) | |
| 39 | load("../res/clfitdf200RDN.Rdata") | |
| 40 | ||
| 41 | lres <- length(res) | |
| 42 | K <- 200 #nrow(clfit$clusinfo) | |
| 43 | rm(res) | |
| 44 | ||
| 45 | ## 2. Process the large file #### | |
| 46 | ||
| 47 | close(con) | |
| 48 | con <- file("~/tmp/2009_full.txt") # Establish a connection to the file | |
| 49 | open(con, "r") # Open the connection | |
| 50 | ||
| 51 | for(b in seq_along(blocks)){ # Reading loop | |
| 52 | nb <- blocks[b] | |
| 53 | actual <- readLines(con = con, n = nb ) | |
| 54 | actual_split <- strsplit(actual, " ") | |
| 55 | rm(actual) | |
| 56 | ||
| 57 | #auxmat <- matrix(unlist(strsplit(actual, " ")), ncol = p + 1, byrow = TRUE) | |
| 58 | #datamat <- t(apply(auxmat[, -1], 1, as.numeric)) # the NA introduced by | |
| 59 | # as.numeric | |
| 60 | auxlist <- lapply(actual_split, function(x) as.numeric(x[-1])) | |
| 61 | datamat <- matrix(unlist(auxlist), ncol = p, byrow = TRUE) | |
| 62 | rm(auxlist) | |
| 63 | ||
| 64 | # are NA strings | |
| 65 | #rownames(datamat) <- substr(auxmat[, 1], 2, 7) | |
| 66 | auxnames <- unlist(lapply(strsplit(actual, " "), "[", 1)) | |
| 67 | rownames(datamat) <- substr(auxnames, 2, 7) | |
| 68 | ||
| 69 | rm(auxmat, actual_split) | |
| 70 | ||
| 71 | synchros <- lapply(resRDN, | |
| 72 | function(ll) { | |
| 73 | aux <- matrix(0, ncol = p, nrow = K) | |
| 74 | for(k in 1:K) { | |
| 75 | clustk <- which(ll == k) | |
| 76 | if(length(clustk) > 1) { | |
| 77 | aux[k, ] <- colSums(datamat[clustk, ]) | |
| 78 | } else { | |
| 79 | aux[k, ] <- datamat[clustk, ] | |
| 80 | } | |
| 81 | } | |
| 82 | aux | |
| 83 | }) | |
| 84 | } | |
| 85 | ||
| 86 | close(con) # close connection to the file | |
| 87 | ||
| 88 | # save(synchros, file = "~/tmp/2009synchrosdf200RND") | |
| 89 |