Issue Combining Date and Time Variables into Single Numeric Datetime Variable

[Scooter56]

I have numeric date and time variables that I would like to combine into a single numeric datetime variable. The code below works on some of the data, but not all of it (more specifically it won't calculate a datetime for any row where the time variable has a two digit hour - see example output below Proc SQL code). The SAS log gives two notes: "Invalid year value" and "Invalid argument to function INPUT. Missing values may be generated." I have tried changing the datetime informat of the input statement but that doesn't help. Any suggestions on how to fix this using Proc SQL (I'm using SAS 9.2) would be appreciated.

proc sql ;
   select a.injury_date
	 ,a.injury_time
	 ,input(put(a.injury_date,date9.)||put(a.injury_time,time8.),datetime17.) as injury_datetime format = datetime19.
     from injury_data ;

injury_date	injury_time	injury_datetime
14SEP2006	1:40:00	14SEP2006:01:40:00
29JUN2006	21:02:00	.
02AUG2006	16:31:00	.
02OCT2006	9:44:00	02OCT2006:09:44:00

[RW9]

Use TOD format:

input(cats(put(a.injury_date,date9.),put(a.injury_time,tod8.)),datetime17.) as injury_datetime format = datetime19.

[Scooter56]

Thanks! I copied your code directly from your post, but it didn't work. The new variable injury_datetime had a missing value for all data rows, and I got the same notes in the SAS log.

[ballardw · Accepted solution]

If the variables are actual SAS date and Time valued then you can use

dhms(a.injury_date, 0,0,a.injury_time) as injury_datetime