About kt_080517

kt_080517 · ‎11-14-2018

Thank you for this! I ran the code - and I think the one issue I have is with the flags for in which there are two or tests within the 6-12 mo period. I want to only have one flag (for the first instance that a test falls within the 6-12 months). The code you provided adds multiple flags.. So for example say I have this additional client, your code would flag both tests (1/13/2016 AND 1/19/2016), I'd only want to flag the 1/13/2016 record. 9 12 2/5/2015 39 23 1/13/2016 39 34 1/19/2016 39 45 2/24/2016 3

kt_080517 · ‎11-13-2018

okay - updated the data!

kt_080517 · ‎11-13-2018

Person Test Test_date status 1 12 4/8/2015 3 1 23 7/7/2015 3 1 34 4/28/2016 3 1 45 10/21/2016 3 1 56 10/21/2016 3 2 22 9/1/2016 3 2 33 10/1/2016 3 2 44 3/1/2017 2 2 55 8/1/2017 3 3 88 10/1/2015 3 3 99 11/1/2015 2 3 76 5/1/2016 3 4 87 5/10/2015 3 4 86 11/10/2015 3 4 85 11/20/2015 3 4 84 1/5/2016 3 4 83 2/16/2016 3 4 82 3/1/2016 2 4 81 4/1/2017 3 I want to flag a change in status - I want to flag the first change that happens in a span of 6-12 months. If there is no change in status, I will keep the second test greater than or equal to 6 months. Essentially, I want to keep 2 records per person - the first ever record and the second record that reports a change (or no change). Expected flags: Person Test Test_date status flag 1 12 4/8/2015 3 . 1 23 7/7/2015 3 . 1 34 4/28/2016 3 1 1 45 10/21/2016 3 . 1 56 10/21/2016 3 . 2 22 9/1/2016 3 . 2 33 10/1/2016 3 . 2 44 3/1/2017 2 1 2 55 8/1/2017 3 . 3 88 10/1/2015 3 . 3 99 11/1/2015 2 . 3 76 5/1/2016 3 1 4 87 5/10/2015 3 . 4 86 11/10/2015 3 4 85 11/20/2015 3 . 4 84 1/5/2016 3 . 4 83 2/16/2016 3 . 4 82 3/1/2016 2 1 4 81 4/1/2017 3 .

kt_080517 · ‎11-02-2018

Can Proc SQL be used here? to bring a value from obs other than the first into a column of its own so all data is in one row?

kt_080517 · ‎11-02-2018

in my current dataset, yes - there are only 2 observations per client. But, I may expand that to multiple obs per client (not exceeding 25 obs)

kt_080517 · ‎11-02-2018

This is my dataset: Client Gender Test_ID Test_Date A F 123 5/1/2015 A F 567 11/1/2015 B M 789 1/1/2017 B M 548 6/1/2017 C M 121 8/1/2016 C M 540 2/1/2017 I want this dataset: Client Gender Test_ID_one Test_ID_two Test_Date_one Test_date_two A F 123 567 5/1/2015 11/1/2015 B M 789 548 1/1/2017 6/1/2017 C M 121 540 8/1/2016 2/1/2017 I think this is a transpose? Any help would be great!

kt_080517 · ‎10-02-2018

The first form I want is the earliest form for a subject, and then I would take the next form that comes after 6 months. For example, say a subject has 3 forms entered on the following dates: 01/01/2018, 03/01/2018, and 08/01/2018. The first form would be dated 01/01/2018, and then the next form would be the one that comes right after a 6 month window, so the form entered on 08/01/2018. Make sense?

kt_080517 · ‎10-02-2018

I would actually just use formdate NOT lag_formdate to determine the 6 months interval

kt_080517 · ‎10-02-2018

SubjectID FormDate Status Lag_formdate interval_months form_num 1 3/24/2015 2 . . 1 1 4/14/2016 3 7/7/2015 9 3 3 8/12/2015 3 . . 1 3 4/5/2016 3 11/17/2015 5 3 5 1/16/2015 2 . . 1 5 9/2/2015 2 2/13/2015 7 3 7 10/15/2015 1 . . 1 7 4/14/2016 3 10/15/2015 6 2 8 4/6/2015 3 . . 1 8 1/21/2016 3 4/6/2015 9 2

kt_080517 · ‎10-02-2018

SubjectID FormDate Status Lag_formdate interval_months form_num 1 3/24/2015 2 . . 1 1 7/7/2015 3 3/24/2015 4 2 1 4/14/2016 3 7/7/2015 9 3 1 7/7/2016 3 4/14/2016 3 4 1 10/12/2016 3 7/7/2016 3 5 3 8/12/2015 3 . . 1 3 11/17/2015 3 8/12/2015 3 2 3 4/5/2016 3 11/17/2015 5 3 3 5/12/2016 3 4/5/2016 1 4 5 1/16/2015 2 . . 1 5 2/13/2015 2 1/16/2015 1 2 5 9/2/2015 2 2/13/2015 7 3 5 5/20/2016 3 9/2/2015 8 4 5 12/9/2016 3 5/20/2016 7 5 7 10/15/2015 1 . . 1 7 4/14/2016 3 10/15/2015 6 2 7 10/27/2016 3 4/14/2016 6 3 8 4/6/2015 3 . . 1 8 1/21/2016 3 4/6/2015 9 2 8 8/23/2016 3 1/21/2016 7 3 Dataset A is above. I want to keep the first form per subject and the first form that comes <= 6 months; so there are 2 forms per client in the output dataset. I have calculated the lag form date and presented in a column above. I would use the lag date minus the form date to get the first assessment after 6 months. I need help getting the first form after 6 months. data B; set A; by subjectID formdate; if first.subjectID; run;

kt_080517 · ‎09-26-2018

I have this dataset, and I want to flag the instance where the status changes. I want to see at the first time the status changes, what is the number of tests received.

Online Status	Offline
Date Last Visited	‎11-14-2018 10:45 AM

Re: Flagging first change in 6-12 month span

Re: Flagging first change in 6-12 month span

Flagging first change in 6-12 month span

Re: Bring data from second obs of client into first obs so all data is...

Re: Bring data from second obs of client into first obs so all data is...

Bring data from second obs of client into first obs so all data is in ...

Re: keep first obs per id and first obs after 6 months

Re: keep first obs per id and first obs after 6 months

Re: keep first obs per id and first obs after 6 months

keep first obs per id and first obs after 6 months

Re: keep first obs per id and first obs after 6 months

Re: Flagging first change in 6-12 month span

Re: Flagging first change in 6-12 month span

Flagging first change in 6-12 month span

Re: Bring data from second obs of client into first obs so all data is...

Re: Bring data from second obs of client into first obs so all data is...

Bring data from second obs of client into first obs so all data is in ...

Re: keep first obs per id and first obs after 6 months

Re: keep first obs per id and first obs after 6 months

Re: keep first obs per id and first obs after 6 months

keep first obs per id and first obs after 6 months

longitudinal data flagging a change in response for subjects that have...