Data are not missing at least two consecutive years

[shawnchen0321]

Hi, all experts.

I have a sample selection criterion which is data are not missing at least two consecutive years.

I have a sample below.

data have;
  input Panelist Year othervars;
  cards;
1          2017 1
1          2019 1
1          2020 1
2          2017 1
2          2020 1
3          2018 1
3          2020 1
;
run;

I want to be like the code below.

data want;
  input Panelist Year othervars;
  cards;
1          2017 1
1          2019 1
1          2020 1
3          2018 1
3          2020 1
;
run;

Does anyone know how to solve this problem?

Thanks in advance.

[mkeintz · Accepted solution]

You have to pass through each panelist twice - once to find gaps, and the second time to reread and output those with no two-year gaps:

data want (drop=_:);
  set have (in=firstpass)  have (in=secondpass);
  by panelist;

  _gap_found + (firstpass=1 and dif(year)>2);
  if first.panelist then _gap_found=0;

  if secondpass and _gap_found=0;
run;

This assumes that the data are sorted by panelist/year. 

Editted note: the DIF(x) function is the result of   x-LAG(x), except it doesn't generate a "missing values were generated ..." note for the first observation.

[shawnchen0321]

It can work. Thanks a lot.