Delete observations under certain condition

[MFraga]

Dear all,

I need to delete observations for a variable when a certain condition happens. Let's me give you an example of my dataset:

data have;

input id variable1 variable_year;

datalines;

1 0 2000

1 0 2001

1 0 2002

1 1 2003

2 0 2007

2 1 2008

3 0 2003

3 0 2004

3 0 2005

4 1 2006

4 . 2007

5 1 2010

5 . 2011

5 . 2012

6 1 2001

6 . 2002

6 . 2003

6 . 2004

6 . 2005

;

run;

What I want as a result would be something like that:

I need to delete observations for a variable when a certain condition happens. Let's me give you an example of my dataset:

data want;

input id variable1 variable_year;

datalines;

1 0 2000

1 0 2001

1 0 2002

1 1 2003

2 0 2007

2 1 2008

3 0 2003

3 0 2004

3 0 2005

;

run;

Si I want to delete the observations by id variable when variable1 = 1 and variable_year = first row. Does anyone could help ?

I have tried this, but it does not work because I just delete the row for that observation.

data want;

set have;

if variable1 = 1 and variable_year=first.variable_year then delete;

run;

[novinosrin]

Pretty straight forward but what is this?

variable_year = first row.

[MFraga]

It is just the first observation for the variable_year. When the "variable_year" is in the first line for the ID and "variable1" is equal to 1 then I want to delete all the observatins for that "id" that meet my condition.

[novinosrin · Accepted solution]

Assuming your data is sorted by ID year

data have;

input id variable1 variable_year;

datalines;
1 0 2000
1 0 2001
1 0 2002
1 1 2003
2 0 2007
2 1 2008
3 0 2003
3 0 2004
3 0 2005
4 1 2006
4 . 2007
5 1 2010
5 . 2011
5 . 2012
6 1 2001
6 . 2002
6 . 2003
6 . 2004
6 . 2005
;

run;

data want;
set have;
by id;
retain k;
if first.id then k=0;
if first.id and variable1=1 then k=1;
if k ne 1;
drop k;
run;

[MFraga]

This works fine. Many thanks!