missing date by group

[monday89]

I have the following dataset.

What I would like to do is

• if group = 2 and (stop_date_time_1 is NOT MISSING and occurs on or before final_stop_date_time) then keep = 1
• else if group = 2 and (stop_date_time_1 is missing) and start_date_time_1 occurs AFTER final_stop_date_time) then keep = 0

SubjectID	final_stop_date_time	start_date_time_1	stop_date_time_1	group
11	12FEB2020:10:59:00	11FEB2020:15:30:00	11FEB2020:16:08:00	2
11	12FEB2020:10:59:00	11FEB2020:16:08:00	12FEB2020:10:59:00	2
11	12FEB2020:10:59:00	12FEB2020:21:40:00	.	2

Want: 

SubjectID	final_stop_date_time	start_date_time_1	stop_date_time_1	group	keep
11	12FEB2020:10:59:00	11FEB2020:15:30:00	11FEB2020:16:08:00	2	1
11	12FEB2020:10:59:00	11FEB2020:16:08:00	12FEB2020:10:59:00	2	1
11	12FEB2020:10:59:00	12FEB2020:21:40:00	.	2	0

I tried the following and doesn't work

data want; set have; 

if group = '2' and 
(stop_date_time_1 <= final_stop_date_time) then keep = 1 ;
else if 
(group = '2' and ((start_date_time_1 > final_stop_date_time ) and missing(stop_date_time_1)) OR 
(group = '2' and (stop_date_time_2 > final_stop_date_time)) 
then keep = 0;
run;

[ballardw]

Doesn't work is awful vague.

Are there errors in the log?: Post the code and log in a code box opened with the <> to maintain formatting of error messages.

No output? Post any log in a code box.

Unexpected output? Provide input data in the form of data step code pasted into a code box, the actual results and the expected results. Instructions here: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-dat... will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the <> icon or attached as text to show exactly what you have and that we can test code against.

I'm going to go out on a limb and guess that part of the problem is you get Keep=1 for missing values of the Stop_date_time.

if group = '2' and 
not missing(stop_date_time) and
(stop_date_time_1 <= final_stop_date_time) then keep = 1 ;

Missing values are always less than any given value. So you need to explicitly limit that part.

Alternate could be to provide a lower limit instead of the missing function:

if group = '2' and 
( 0 <= stop_date_time_1 <= final_stop_date_time) then keep = 1 ;

but that approach may not be appropriate if you do not know what your minimum actual value of the stop_date_time should be.

I am assuming that your datetime values are SAS datetimes valued. If they are character then any comparison with < or > is almost always a very bad idea. 'January' comes after 'April' in terms of inequalities.

[monday89]

There is no error but this how final dataset looks. The last record should keep = 0 because stop_date_time_2 is null and start_date_time_1 occurs after final_stop_date_time

SubjectID	final_stop_date_time	start_date_time_1	stop_date_time_2	group	keep
11	12FEB2020:10:59:00	11FEB2020:15:30:00	11FEB2020:16:08:00	2	1
11	12FEB2020:10:59:00	11FEB2020:16:08:00	12FEB2020:10:59:00	2	1
11	12FEB2020:10:59:00	12FEB2020:21:40:00	.	2	1

[ballardw]

Count the ( and ) in this:

(group = '2' and ((start_date_time_1 > final_stop_date_time ) and missing(stop_date_time_1)) OR 
(group = '2' and (stop_date_time_2 > final_stop_date_time)) 

you have an unbalanced () somewhere.

Not nice to include variables in code without example data.

Since every single one of your conditions involves group='2' a good style choice is to factor out the common element:

If group='2' then do;
   <condition one>
   <next condition>
   <etc>
end;

which will make it easier to read and understand that this block all involves group='2'. And sometimes you can introduce logic errors with minor typos for repeated code.

[smantha]

data new;
informat final_stop_date_time	start_date_time_1	stop_date_time_1 datetime19.;
input SubjectID	final_stop_date_time	start_date_time_1	stop_date_time_1	group $;
format final_stop_date_time	start_date_time_1	stop_date_time_1 datetime19.;
if group = '2' and 
(stop_date_time_1 <= final_stop_date_time) then keep = 1 ;
if 
(group = '2' and ((start_date_time_1 >= final_stop_date_time ) and missing(stop_date_time_1))) OR 
(group = '2' and (stop_date_time_1 > final_stop_date_time)) 
then keep = 0;
cards;;;
11	12FEB2020:10:59:00	11FEB2020:15:30:00	11FEB2020:16:08:00	2
11	12FEB2020:10:59:00	11FEB2020:16:08:00	12FEB2020:10:59:00	2
11	12FEB2020:10:59:00	12FEB2020:21:40:00	.	2
;;;

run;

reasons why it did not work before:

1. (stop_date_time_2 > final_stop_date_time)  stop_date_time_2 does not exist

2. It did not go into else condition as when the date column is missing first if condition is always true