selecting rows given condition

[DathanMD]

Hello mates;

I am new to SAS and I am stuck on my next step of data manipulation. I am not providing any code because I have no crew where to begin from with this logic.

I have data as shown below. I am interested in only keeping rows when the makes the initial unique change from the previous position. For example staff one first made the first ever change on 02jan2001 (from being cleaner to administrator). For staff 2, the first change ever was on 02jan2019 from Manager to Deputy_Director). For staff who have no change in position (for example staff 3), I would like to just retain the first row.

Data I have;

staff	Start_date	End_date	Job position
1	01jan2000	01jan2001	cleaner
1	02jan2001	01jan2002	administrator
1	02jan2002	01jan2003	Deputy_manager
1	02jan2004	01jan2005	Manager
2	01jan2000	02jan2004	Manager
2	03jan2004	03jan2006	Manager
2	02jan2006	01jan2009	Deputy_Director
2	02jan2009	01jan2012	Director
2	01jan2012	01jan2014	Unemployed
2	02jan2014	01jan2016	Director
3	01jan2000	06jan2000	unemployed
3	07jan2000	01jul2000	Unemployed
4	01jan2000	01jun2000	Cleaner
4	02jun2000	01aug2000	Manager
5	02aug2006	01jan2019	Director

Data I want;

Patient	Start_date	End_date	Job position
1	02jan2001	01jan2002	administrator
2	02jan2006	01jan2009	Deputy_Director
3	01jan2000	02jan2000	unemployed
4	01jan2000	01jan2000	Manager

Thank you.

Dathan Byonanebye

[DathanMD]

Hello mates;

I am new to SAS and I am stuck on my next step of data manipulation. I am not providing any code because I have no crew where to begin from with this logic.

I have data as shown below. I am interested in only keeping rows when the makes the initial unique change from the previous position. For example staff one first made the first ever change on 02jan2001 (from being cleaner to administrator). For staff 2, the first change ever was on 02jan2019 from Manager to Deputy_Director). For staff who have no change in position (for example staff 3), I would like to just retain the first row.

Data I have;

staff	Start_date	End_date	Job_position
1	01jan2000	01jan2001	cleaner
1	02jan2001	01jan2002	administrator
1	02jan2002	01jan2003	Deputy_manager
1	02jan2004	01jan2005	Manager
2	01jan2000	02jan2004	Manager
2	03jan2004	03jan2006	Manager
2	02jan2006	01jan2009	Deputy_Director
2	02jan2009	01jan2012	Director
2	01jan2012	01jan2014	Unemployed
2	02jan2014	01jan2016	Director
3	01jan2000	06jan2000	unemployed
3	07jan2000	01jul2000	Unemployed
4	01jan2000	01jun2000	Cleaner
4	02jun2000	01aug2000	Manager
5	02aug2006	01jan2019	Director

Data I want;

Patient	Start_date	End_date	Job_position
1	02jan2001	01jan2002	administrator
2	02jan2006	01jan2009	Deputy_Director
3	01jan2000	02jan2000	unemployed
4	01jan2000	01jan2000	Manager

Thank you.

Dathan Byonanebye

[andreas_lds]

@DathanMD Please don't double post questions.

[PeterClemmensen]

Why is Staff=5 not included? She has no change in position, though only 1 obs?

[ed_sas_member]

Hi @DathanMD 

In addition to @PeterClemmensen 's question, how would like you manage the case of staff=3?

3	01jan2000	06jan2000	unemployed
3	07jan2000	01jul2000	Unemployed

- option 1: retrieve the first row (-> this is what you described in your initial post)

- option 2: collapse dates so that the end date appears as 01jul2000. In this case how would you manage time interruptions?

[DathanMD]

Hi @PeterClemmensen


Option 2 is perfect pick for me. Thanks

Dathan