modifier spain to include code for bad measurment

merge.data.PARACOU.R

@@ -9,24 +9,29 @@ rm(list = ls()); source("./R/format.function.R"); library(reshape)
 data.paracou <- read.table("./data/raw/DataParacou/20130717_paracou_1984_2012.csv",header=TRUE,stringsAsFactors=FALSE,sep = ";", na.strings = "NULL")
 #barplot(apply(!is.na(data.paracou[,paste("circ_",1984:2012,sep="")]),MARGIN=2,FUN=sum),las=3)
 
-data.paracou <- data.paracou[,c(1:8,61:62,73:74,85:86,96:97)]
-colnames(data.paracou) <- c("forest","plot","subplot","tree","vernacular","taxonid","x","y","circum2001","dead2001","circum2005","dead2005","circum2009","dead2009","yeardied","typedeath")
+
+
+data.paracou <- data.paracou[,c("foret","parcelle","carre","arbre","vernaculaire","idtaxon",
+                                "x","y","circ_2001","code_2001","circ_2005","code_2005",
+                                "circ_2009","code_2009","campagne_mort","type_mort")]
+colnames(data.paracou) <- c("forest","plot","subplot","tree","vernacular","taxonid","x","y","circum2001","code2001","circum2005","code2005","circum2009","code2009","yeardied","typedeath")
 for(k in 7:14) { 
 	data.paracou[,k] <- gsub(",",".",data.paracou[,k]); data.paracou[,k] <- as.numeric(data.paracou[,k]) } ## Replace all , in decimals with .
 data.paracou$treeid <- apply(data.paracou[,1:4],1,paste,collapse="."); ## Create a tree id
 data.paracou <- data.paracou[,c(ncol(data.paracou),1:(ncol(data.paracou)-1))]	
 
+## keep only tree alive in 2001
+data.paracou <- subset(data.paracou,subset=!(as.numeric(data.paracou[["yeardied"]])<=2001 & !is.na(data.paracou[["yeardied"]])))
+
 ### read species names
 species.clean <- read.csv("./data/raw/DataParacou/20130717_paracou_taxonomie.csv",stringsAsFactors=FALSE, header = T, sep = ";")
 
-## Create a species code to try and species.clean and data.paracou - NEED TO CHECK WHETHER SPCODE COMES FROM TAXONID OR VERNACULAR OR BOTH
-species.clean$sp <- apply(species.clean[,1:5],1,paste,collapse=".")
-data.paracou$sp <- apply(data.paracou[,3:7],1,paste,collapse=".")
-data.paracou <- merge(data.paracou, as.data.frame(species.clean[,c(10,13)]), by = "sp", sort = F)
-## "Simplify" the species.code to just vernacular and taxonid, which is hopefully sufficient to identify species
-species.clean$sp = apply(species.clean[,4:5],1,paste,collapse=".")
-data.paracou$sp = apply(data.paracou[,7:8],1,paste,collapse=".")
-length(unique(data.paracou$sp))
+## SPECIES CODE COME FROM idTaxon in paracou_taxonomie and taxonid in paracou_1984_2012 to match the traits data we need to use the "Genus species"
+## we better work not work with vernacular because this doesn't match necesseraly the Genus species taxonomie
+
+species.clean$sp <- species.clean[["idTaxon"]]
+data.paracou$sp <- data.paracou[["taxonid"]]
+## data.paracou <- merge(data.paracou, as.data.frame(species.clean[!duplicated(species.clean[["sp"]]),c("Genre","Espece","sp")]), by = "sp", sort = FALSE)
 
 ######################################
 ## MASSAGE TRAIT DATA
@@ -39,15 +44,23 @@ length(unique(data.paracou$sp))
 data.paracou2 <- data.paracou[rep(1:nrow(data.paracou),each=2),c(1:10,(ncol(data.paracou)-2):ncol(data.paracou))]
 rownames(data.paracou2) <- 1:nrow(data.paracou2); data.paracou2 <- as.data.frame(data.paracou2)
 data.paracou2$yr1 <- rep(c(2001,2001+4),nrow(data.paracou)); data.paracou2$yr2 <- rep(c(2005,2005+4),nrow(data.paracou))
-data.paracou2$year <- rep(c(5,5),nrow(data.paracou))
+data.paracou2$year <- rep(c(4,4),nrow(data.paracou))
 data.paracou2$dbh1 <- c(rbind(data.paracou$circum2001/pi,data.paracou$circum2005/pi))
 data.paracou2$dbh2 <- c(rbind(data.paracou$circum2005/pi,data.paracou$circum2009/pi))
-data.paracou2$dead <- c(as.numeric(rbind(data.paracou$dead2005 > 0,data.paracou$dead2009 > 0)))
-	
+data.paracou2$code1 <- c(as.numeric(rbind(data.paracou$code2001,data.paracou$code2005)))
+data.paracou2$code2 <- c(as.numeric(rbind(data.paracou$code2005,data.paracou$code2009)))
+data.paracou2$dead <- rep(0,nrow(data.paracou)*2)
+data.paracou2$dead[c(as.numeric(data.paracou[["yeardied"]]) %in% 2001:2005 & (!is.na(data.paracou[["yeardied"]])),
+                     as.numeric(data.paracou[["yeardied"]]) %in% 2006:2009 & (!is.na(data.paracou[["yeardied"]])))] <- 1
+
+
 data.paracou <- data.paracou2	
 
 ## change unit and names of variables to be the same in all data for the tree 
 data.paracou$G <- 10*(data.paracou$dbh2-data.paracou$dbh1)/data.paracou$year ## diameter growth in mm per year
+data.paracou$G[data.paracou$code1>0] <- NA
+data.paracou$G[data.paracou$code2>0] <- NA
+
 data.paracou[which(data.paracou$G < -50),] ## THERE SEEMS TO BE SOME PROBLEMS WITH THE DBH DATA
 data.paracou$D <- data.paracou[["dbh1"]]; data.paracou$D[data.paracou$D == 0] <- NA ;## diameter in cm
 data.paracou$plot <- apply(data.paracou[,c("forest","plot","subplot")],1,paste,collapse=".") ## plot code

merge.data.SPAIN.R

@@ -44,7 +44,7 @@ write.csv(data.max.height,file="./data/process/data.max.height.spain.csv") # I w
 
 ## change unit and names of variables to be the same in all data for the tree 
 data.spain$G <- data.spain[["adbh"]] ## diameter growth in mm per year
-data.spain$year <- rep(NA,length(data.spain[["adbh"]])) ## number of year between measurement - MISSING
+data.spain$year <- data.spain[["years"]] ## number of year between measurement - MISSING
 data.spain$D <- data.spain[["dbh1"]]/10 ## diameter in mm convert to cm
 data.spain$dead <- as.numeric(data.spain[["Life_status"]] == "dead") ## dummy variable for dead tree 0 alive 1 dead - MIGHT WANT TO CHANGE THIS TO BE BASED ON MORTALITY_CUT
 data.spain$sp <- as.character(data.spain[["SP_code"]]) ## species code