Why are some datetime values being deleted when I round the time in R? -

i importing series of csv files r. these contain date/time column, id, , 2 columns of temperature values.

this give example of data looks like:

id<-c(1,2,3,4) date.time<-as.character(c("12/03/17 00:21:28", "12/03/17 02:21:28", "12/03/17 04:21:28", "12/03/17 06:21:28")) temp1<-c(-3.568,-3.568,-3.598,-3.598) temp2<-c(-11.577,-11.577,-11.541,-11.433) df<-data.frame(id,date.time,temp1,temp2) 

because date/time not in format want, i've been using strptime , formatting them posixlt.

like:

df$date.time<-strptime(df$date.time, "%d/%m/%y %h:%m:%s") df$date.time<- as.posixlt(df$date.time, "%y/%m/%d %h:%m:%s", tz="gmt0") 

this works fine , gives data looks like:

id  date.time            temp1   temp2 1   2017-03-12 0:21:28  -3.568  -11.577 2   2017-03-12 2:21:28  -3.568  -11.577 3   2017-03-12 4:21:28  -3.598  -11.541 4   2017-03-12 6:21:28  -3.598  -11.433 

however, want round time part of date.time column nearest hour. i've been using:

df$date.time<-round(df$date.time, units="hours") 

this works fine 99% of time. however, in of files, r deleting date.time values, seemingly @ random, , giving na. 1 or 2 values in each file being deleted , can't see reason these particular values deleted. example:

id  date.time         temp1   temp2 1   2017-03-12 0:00:00  -3.568  -11.577 2   na                  -3.568  -11.577 3   2017-03-12 4:00:00  -3.598  -11.541 4   2017-03-12 6:00:00  -3.598  -11.433 

from i've read, date/time values can finicky seems oddly random.

does know might causing issue , if there better way round time part of posixlt values?

update: seems times deleted @ 2am on march 12th. so, many of times should rounded 2017-03-12 02:00:00 being replaced nas. not happen of csv files, half. why r having problems reading particular date?

thanks!

not adding timezone strptime corrupting string.

lets @ head of df$date.time. missing/different time zones.

 head(df$date.time) [1] "2017-03-12 00:21:28 pst" [2] "2017-03-12 02:21:28"     [3] "2017-03-12 04:21:28 pdt" [4] "2017-03-12 06:21:28 pdt" 

by adding tz="gmt0" strptime function should desired result.

df$date.time<-strptime(df$date.time, "%d/%m/%y %h:%m:%s", tz="gmt0") df$date.time<- as.posixlt(df$date.time, "%y/%m/%d %h:%m:%s", tz="gmt0") df$date.time<-round(df$date.time, units="hours")  > head(df$date.time) [1] "2017-03-12 00:00:00 gmt" [2] "2017-03-12 02:00:00 gmt" [3] "2017-03-12 04:00:00 gmt" [4] "2017-03-12 06:00:00 gmt"