re: Counting Number of Days Error

[twildone]

Hi, I seem to be getting an error in the calculation of the number of days between two dates. For example, in my dataset, I have 20130220 and 20130415 and using the INTCK Function, I am getting -596 where it should be 55. Not too sure why or where the negative number is coming from. Thanks in advance.

DATA SUMMARY;

SET SUMMARY;

IF (ID = LAG(ID)) AND (CLAIM_DATE >= LAG(CLAIM_DATE)) THEN DO;

DIFF = INTCK('DAY',INPUT(PUT(LAG(CLAIM_DATE),8.),yymmdd8.),INPUT(PUT(CLAIM_DATE,8.),yymmdd8.))+1;

END;

RUN;

[RW9]

Could you provide some test data there.  Not sure why you are putting lag(claim_date) to number then reading in as date. 

[LinusH]

Try to isolate your problem, i.e. separating your LAG(), input() and intck() into separate statements.

Use put _ALL_ to monitor the results of your operations.

But I have never really understood the value of using INTCK without arguments for straight dates. You just use simple subtraction on SAS dates....?

[Loko]

hello,

probably it is because of LAG(CLAIM_DATE), executed conditionally. if sometimes the condition is not true the results are not the expected ones.

[twildone]

Hi.....here is some output when I ran the code.

ID	CLAIM_DATE	DIFF
100003976	20130709
100003976	20130808	37
100003976	20130909	33
100003976	20131007	29
100003976	20131104	29
100003976	20131114	11
100004028	20130220
100004028	20130415	-596
100004028	20130622	69
100004028	20130819	59
100004028	20131016	59
100004028	20131231	77
100004028	20140224	56
100009987	20120222
100009987	20120319	-951
100009987	20120420	33
100009987	20120519	30
100009987	20120618	31
100009987	20120917	92
100009987	20121219	94
100009987	20130328	100

[twildone]

Thanks everyone. Apparently, it looks like the INTCK Function does not like the LAG Function as part of the Start or End Date. I created another variable for the Lag claim date and then used the new variable with the INTCK Function and it seems to work fine now.....Thanks Once Again.