I'm trying to save some time information included data with write_csv.

But it keeps showing T and Z character (ISO8601 format as I know).

Forexample, 2022-12-12 08:00:00 is shown as 2022-12-12T08:00:00Z on csv file open with notepad

I want keep original data format after saving csv file, but I couldnt find option for this.

Just saw some article but there are no answer.

Other Q

CodePudding user response：

Here are two solutions.

First write a data set to a temp file.

library(readr)

df1 <- data.frame(datetime = as.POSIXct("2022-12-12 08:00:00"),
                  x = 1L, y = 2)
csvfile <- tempfile(fileext = ".csv")

# write the data, this is the problem instruction
write_csv(df1, file = csvfile)

^{Created on 2023-01-31 with reprex v2.0.2}

1. Change nothing

This is probably not what you want but read_csv recognizes write_csv's ISO8601 output format, so if the data is written to file with write_csv and read in from disk with read_csv the problem doesn't occur.

# read from file as text, problem format is present
readLines(csvfile)
#> [1] "datetime,x,y"             "2022-12-12T08:00:00Z,1,2"

# read from file as spec_tbl_df, problem format is not present
read_csv(csvfile, show_col_types = FALSE)
#> # A tibble: 1 × 3
#>   datetime                x     y
#>   <dttm>              <dbl> <dbl>
#> 1 2022-12-12 08:00:00     1     2

^{Created on 2023-01-31 with reprex v2.0.2}

2. Coerce to "character"

If the datetime column of class "POSIXct" is coerced to character the ISO8601 format is gone and everything is OK. And afterwards read_csv will recognize the datetime column.
This is done in a pipe, below with the base pipe operator introduced in R 4.1, in order not to change the original data.

# coerce the problem column to character and write to file
# done in a pipe it won't alter the original data set
df1 |>
  dplyr::mutate(datetime = as.character(datetime)) |>
  write_csv(file = csvfile)

# check result, both are OK
readLines(csvfile)
#> [1] "datetime,x,y"            "2022-12-12 08:00:00,1,2"

read_csv(csvfile, show_col_types = FALSE)
#> # A tibble: 1 × 3
#>   datetime                x     y
#>   <dttm>              <dbl> <dbl>
#> 1 2022-12-12 08:00:00     1     2

^{Created on 2023-01-31 with reprex v2.0.2}

Final clean up.

unlink(csvfile)