Delete a matched control if the corresponding matched case is deleted

[arzon]

Hi Dears,

I would greatly appreciate your help to delete some matched observations from a dataset. Briefly, I have a dataset in which cases are matched with controls and then a variable (exclusion_variable) is used to exclude ID that met the criterion of exclusion. As presented in the table below, I would like to delete the corresponding controls as the case (ID=1x) is to be excluded. 

ID    matching_ID   case   exclusion_variable

1x     aa                  1         1

2x     aa                  0         0

3x     aa                  0         0

Thank you very match for your valuable help!

[Reeza]

@arzon welcome to the SAS forums. 

Unfortunately your question is unclear. Can you please clarify what you have and what you want as output?

---

@arzon wrote:

Hi Dears,

I would greatly appreciate your help to delete some matched observations from a dataset. Briefly, I have a dataset in which cases are matched with controls and then a variable (exclusion_variable) is used to exclude ID that met the criterion of exclusion. As presented in the table below, I would like to delete the corresponding controls as the case (ID=1x) is to be excluded. 

ID    matching_ID   case   exclusion_variable

1x     aa                  1         1

2x     aa                  0         0

3x     aa                  0         0

Thank you very match for your valuable help!

---

[arzon]

Hi Reeza,

Thanks for your reply and sorry for being unclear (I am not fully fluent in English and that may explain why)

To clarify, lets consider the 'exclusion_variable' as 'death' and that the ID=1x  is dead (case=1) and is to be excluded from the dataset (if death=1 then delete). As the ID 2x, and  3x are matched to the ID 1x (i.e. same matching_ID) I would like to delete these ID from the dataset to preserve the matching nature of the overall dataset.  So, How can I can suppress these matched ID?

Hope this clarifies.

Thank you very much

[Reeza]

So based on your sample data what do you expect as output?

[arzon]

ID    matching_ID   case   exclusion_variable

1x     aa                  1         1

2x     aa                  0         0

3x     aa                  0         0

4x     bb                  1         0

5x     bb                  0         0

6x     bb                  0         0

From the above table, my output will be:

ID    matching_ID   case   exclusion_variable

4x     bb                  1         0

5x     bb                  0         0

6x     bb                  0         0

[mkeintz · Accepted solution]

Why do your refer BOTH to variable CASE and variable EXCLUSION_VAR?  Do you just mean to filter out all matching_id's for which any record has an EXCLUSION_VAR=1?

Edited addition:

If so, then this will do:

data have;
  input id :$2. matching_id :$2. case exclusion_variable;
datalines;
1x     aa                  1         1
2x     aa                  0         0
3x     aa                  0         0
4X     bb                  0         0
run;

data want (drop=rc);
  set have;
  if _n_=1 then do;
    declare hash exclude_id (dataset:'have (keep=matching_id exclusion_variable where=(exclusion_variable=1))');
    exclude_id.definekey('matching_id');
    exclude_id.definedone();
  end;
  rc=exclude_id.check();
  if rc^=0;
run;
        

[arzon]

Hi  mkeintz,

You exactly provided the solution to my query.

Thank you very much for your support.