how to pick "Y" from a variable if they are multiple observations

[vraj1]

is there any way to pick "Y" from a variable with same id, 

I have a id variable and flag variable with multiple observations

101  Y

101   N

101  N

102 N

102 N

103 Y

is there a way where we can pick 101 and only "Y" from it and ignore "N" and in second id only take one record "N" and omitt the duplicate.

[PeterClemmensen]

Like this?

data have;
input id flag $;
datalines;
101 Y
101 N
101 N
102 N
102 N
103 Y
;

proc sort data=have;
	by id flag;
run;

data want;
	set have;
	if last.id;
run;

[ballardw]

If an ID has multiple Y flags what do you want for output? All N flags but more than 2 (your description does not actually cover this case as only mention similar for the second id. What if the 5th id has 4 flags?)

Best to provide an example of what the actual output should look like for given example input.

[novinosrin]

+1 @ballardw  Sir. I was gonna ask that question as I was wary too. You beat me.

Hello @vraj1   is there  a trick like,

Pick all Y's if multiples with an N or not,  and only distinct N if all N's?