CODEWITHSUNDEEP
rposixct
Fomalhaut -C
Fomalhaut -C

Reputation: 320

Can't convert to POSIXct (R)

Given a data set "eur_usd" of class data.frame:

     pair                    dt     bid     ask
1 EUR/USD 20180401 21:02:24.820 1.23178 1.23286
2 EUR/USD 20180401 21:02:25.304 1.23156 1.23285
3 EUR/USD 20180401 21:02:25.358 1.23155 1.23285

and with

> class(eur_usd$dt)
[1] "character"

I can't coerce $dt to POSIXct - everything is just NA:

strptime(eur_usd$dt, "%Y%m%d %H:%M:%S.%f")
   [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
  [36] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

as.POSIXct(eur_usd$dt, format = "%Y%m%d %H:%M:%S.%f", tz = "GMT")
   [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

Am I doing something wrong?

Upvotes: 0

Views: 116

Answers (2)

Dirk is no longer here
Dirk is no longer here

Reputation: 368241

@Rich showed you how to correctly specify the required format when you use a converter requiring a format.

Alternatively, you could use the anytime package which does not need a format specififier (but heuristically tries a number of plausible alternatives):

R> anytime::anytime(c("20180401 21:02:24.820", "20180401 21:02:25.304",
+                   "20180401 21:02:25.358"))
[1] "2018-04-01 21:02:24.819 CDT" "2018-04-01 21:02:25.303 CDT" 
[3] "2018-04-01 21:02:25.358 CDT"     
R>

Upvotes: 1

Rich Scriven
Rich Scriven

Reputation: 99331

You should use %OS for milliseconds

as.POSIXct(df$dt, format = "%Y%m%d %H:%M:%OS")
# [1] "2018-04-01 21:02:24 PDT" "2018-04-01 21:02:25 PDT" 
# [3] "2018-04-01 21:02:25 PDT"

Upvotes: 2

Related Questions