David, veja se te ajuda.

library(R.utils)

lines <- "

5928671,14,-3.831290,-38.50146,2015-02-01 00:00:01

5926098,44,-3.709390,-38.58986,2015-02-02 00:01:13

5915217,145,-3.743671,-38.48788,2015-02-03 00:01:16

5444641,174,-3.766271,-38.57452,2015-02-01 00:01:38

2250561,977,-3.738654,-38.48400,2015-02-02 00:01:47

5671754,14,-3.831290,-38.50146,2015-02-03 00:02:43

con <- textConnection(lines)

df <- read.csv(con, header = F)

close(con)

dias <- unique(as.Date(df[, 5]))

for (i in 1:length(dias)) {

dfi <- df[as.Date(df[, 5]) == dias[i], ]

diretorio <- paste(getwd(), "rsc/data", format(dias[i], format = "%Y-%m-%d"),

sep = "/", collapse = "")

mkdirs(diretorio)

write.csv(dfi, paste(diretorio, "saida.csv", sep = "/", collapse = ""))

}