Home/ Questions/Q 9258645
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-18T12:27:36+00:00

I am working on a dataframe which I have previously intergrated the time and

  • 0

I am working on a dataframe which I have previously intergrated the time and date into one column (called timestamp):

a <-c(1:21)
D <- c("2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14", "2012/12/14")
Time <- c("18:40:37", "18:40:48", "18:40:58", "18:41:08","18:41:18","18:41:28","18:41:38","18:41:48","18:41:58","18:42:08","18:42:18","18:42:28","18:42:38","18:42:48","18:42:58","18:43:08","18:43:18","18:42:28", "18:44:18", "18:44:28", "18:44:28")
df1 <- data.frame(a, D, Time)
df1 <- within(df1, { timestamp=format(as.POSIXct(paste(D, Time)), "%d/%m/%Y %H:%M:%S") })

How would I subset the dataframe to exclude values after a specific point in time? I found some code in Stackoverflow for a similar question that I thought might be able to help, but I am struggling to get the time element to work:

subset(df1, format.Date(timestamp, ""%d/%m/%Y %H:%M:%S"") >"14/12/2012 18:42:00")

Any advice would be very much appreciated.

Edit:
I am struggling to get the code detailed below working on my real data. A dput() of the first four rows of my dataframe are listed at the end of this post. I previously used the line of code recommended by @Arun to timestamp my data.

gps <- within(gps, { timestamp=format(as.POSIXct(paste(LOCAL.DATE, LOCAL.TIME)), 
+                                       "%d/%m/%Y %H:%M:%S") })

If I try and apply the second part of the code (strptime…) I get the error message:
Error in $<-.data.frame(*tmp*, “timestamp”, value = list(sec = c(37, :
replacement has 30208 rows, data has 4
This sort of explains when I try and appy the code to the whole of my data I get 8 rows of many numbers, separated by a comma. If you can help me in any way, I would be extremely grateful.

structure(list(timestamp = c("14/12/2012 18:40:37", "14/12/2012 18:40:48", 
"14/12/2012 18:40:58", "14/12/2012 18:41:08"), LATITUDE = c(54.77769505, 
54.77765729, 54.77768751, 54.7777021), LONGITUDE = c(-1.56627049, 
-1.56639255, -1.56626555, -1.56662523), HEIGHT = c(" 173.911 M", 
" 161.742 M", " 146.905 M", " 138.016 M"), SPEED = c(" 0.465 km/h", 
" 0.728 km/h", " 4.574 km/h", " 17.335 km/h")), .Names = c("timestamp", 
"LATITUDE", "LONGITUDE", "HEIGHT", "SPEED"), row.names = c(NA, 
4L), class = "data.frame")

Second edit: Many thanks @Arun for the solution. I was a bit confused how I was suppose to use the code as my data is originally in date and time columns (LOCAL.DATE and LOCAL.TIME). So I used the first line of code from your orginal solution, and then the second line from your revised edits.

This is the code I used:

gps <- within(gps, { timestamp=format(as.POSIXct(paste(LOCAL.DATE, LOCAL.TIME)), 
                                      "%d/%m/%Y %H:%M:%S") })

gps$timestamp <- strptime(gps$timestamp, "%Y-%m-%d %H:%M:%S")

However now I get a string of NAs (and some -1s). Apologies if I used the code in the incorrect way…

Third edit
Apologies for the confusion @Arun. When I try it both ways round for the date column, I get errors. If I keep it as yr/m/d, how the original data was formatted, I get dput() of :

structure(list(timestamp = c("2012/12/14 18:40:37", "2012/12/14 18:40:48", 
"2012/12/14 18:40:58", "2012/12/14 18:41:08"), LATITUDE = c(54.77769505, 
54.77765729, 54.77768751, 54.7777021), LONGITUDE = c(-1.56627049, 
-1.56639255, -1.56626555, -1.56662523), HEIGHT = c(" 173.911 M", 
" 161.742 M", " 146.905 M", " 138.016 M"), SPEED = c(" 0.465 km/h", 
" 0.728 km/h", " 4.574 km/h", " 17.335 km/h")), .Names = c("timestamp", 
"LATITUDE", "LONGITUDE", "HEIGHT", "SPEED"), row.names = c(NA, 
4L), class = "data.frame")

If I then use:

gps2$timestamp <- strptime(gps2$timestamp, "%Y/%m/%d %H:%M:%S")

… and try to view the dataframe in R Studio’s workspace window, the R session aborts.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    Its better to load character vectors as such and not factors using stringsAsFactors = FALSE (as shown below)

    # make sure character columns are not converted to factors
df1 <- data.frame(a, D, Time, stringsAsFactors = FALSE)

    Then,

    df1 <- within(df1, { timestamp=format(as.POSIXct(paste(D, Time)), 
                               "%d/%m/%Y %H:%M:%S") })
# convert timestamp here
df1$timestamp <- strptime(df1$timestamp, "%d/%m/%Y %H:%M:%S")

    Now, try subset this way:

    # now subset
subset(df1, timestamp > strptime("14/12/2012 18:42:00", "%d/%m/%Y %H:%M:%S"))


#     a          D     Time           timestamp
# 10 10 2012/12/14 18:42:08 2012-12-14 18:42:08
# 11 11 2012/12/14 18:42:18 2012-12-14 18:42:18
# 12 12 2012/12/14 18:42:28 2012-12-14 18:42:28
# 13 13 2012/12/14 18:42:38 2012-12-14 18:42:38
# 14 14 2012/12/14 18:42:48 2012-12-14 18:42:48
# 15 15 2012/12/14 18:42:58 2012-12-14 18:42:58
# 16 16 2012/12/14 18:43:08 2012-12-14 18:43:08
# 17 17 2012/12/14 18:43:18 2012-12-14 18:43:18
# 18 18 2012/12/14 18:42:28 2012-12-14 18:42:28
# 19 19 2012/12/14 18:44:18 2012-12-14 18:44:18
# 20 20 2012/12/14 18:44:28 2012-12-14 18:44:28
# 21 21 2012/12/14 18:44:28 2012-12-14 18:44:28

    Edit: Try this:

    df1 <- within(df1, { timestamp=as.POSIXct(timestamp, format = "%d/%m/%Y %H:%M:%S") })
df1$timestamp <- strptime(df1$timestamp, "%Y-%m-%d %H:%M:%S")
    • 0