Store data according to starting date - r

Question

I am working with multiple csv files stored in a folder (2488). Each csv contains monthly data from gaugin stations with the format years * months. I want to store all these csv files in a big data frame in which cols are the IDs of the different gaugin stations and the rows the times %Y-%m-%d.
For this purpose, I have list all the files with list.files:
a <- list.files(pattern="./*.csv",full.names=TRUE)
And create a dummy data frame with the final dimensions:
gst <- data.frame(NA,1860,2488) # 1860 times - 2488 stations
Each csv file starts in a different date. The earliest date found is January 1863, so I created the data frame with 1860 rows from that starting date to June 2017.
I create a sequence of dates to name the rows of gst:
s <- paste0(1863,"-01-01")
e <- paste0(2017,"-12-31")
ss<- chron(s, format='y-m-d')
ee<- chron(e, format='y-m-d')
dates <- seq.dates(ss,ee,by='months')
In the following loop, I read each csv file. First I change the initial data frame format: years * months + total column.
# Initial format
Jan Feb Mar Apr
1993 NA 0.05 0.05 0.06
1994 0.18 0.15 0.1 0.19
1995 0.22 0.23 0.26 0.11
1996 0.14 0.11 0.1 0.08
1997 0.12 0.16 0.07 0.05
1998 0.12 0.07 0.12 0.18
1999 0.07 0.32 0.14 0.15
2000 0.13 0.22 0.15 0.1
2001 0.18 0.09 0.5 0.26
to a single column data frame (kk.df) with data stored as:
Date Value
93-01-01 NA
93-02-01 0.05
93-03-01 0.05
93-04-01 0.06
93-05-01 0.05
93-06-01 0.05
93-07-01 0.03
93-08-01 0.03
93-09-01 0.05
93-10-01 0.09
93-11-01 0.04
93-12-01 0.10
This is the loop I am working with:
for (i in 1:length(a)){
kk <- read.csv(a[i])
colnames(kk) <- c(seq(1,12,1),'total') # 12 (months) columns and a total column
kk.ts <- ts(as.vector(t(as.matrix(kk))),
start=as.numeric(c(rownames(kk)[1],1)), end= as.numeric(c(rownames(kk)[dim(kk)[1]],12)),frequency=12)
kk.df <- as.data.frame(kk.ts)
colnames(kk.df) <- a[[i]]
a <- paste0(start,"-01-01")
b <- paste0(end,"-12-31")
ac<- chron(a, format='y-m-d')
bc<- chron(b, format='y-m-d')
times <- seq.dates(ac,bc, by="months")
rownames(kk.df) <- times
gst[i,] <- kk.df
}
My question is, as I am storing the same number of columns as gaugin stations I have (2488 stations) and each station start at a different Year-month, how can I specify when store each i for each station, the row in which it must start?
If i = 1 and the first record is in 1993-01-01, I want that column to start at the row of gst that corresponds to 1993-01-01 and so on with the rest of the stations.
Thank you so much.

Answer 1

Maybe you could incorporate a join in your for loop for that:
df = data.frame(date = seq(Sys.Date(),Sys.Date()+3,by=1))
station1 = data.frame(date = seq(Sys.Date()+2,Sys.Date()+3,by=1),data = c(1,2))
station2 = data.frame(date = seq(Sys.Date()+1,Sys.Date()+2,by=1),data = c(2,3))
df = df %>% left_join(station1) %>% rename(station1=data)
df = df %>% left_join(station2) %>% rename(station2=data)
Input:
> df
date
1 2017-07-17
2 2017-07-18
3 2017-07-19
4 2017-07-20
> station1
date data
1 2017-07-19 1
2 2017-07-20 2
> station2
date data
1 2017-07-18 2
2 2017-07-19 3
Output:
> df
date station1 station2
1 2017-07-17 NA NA
2 2017-07-18 NA 2
3 2017-07-19 1 3
4 2017-07-20 2 NA