Help with filtering logic...

[NKormanik]

Given A, B, C

In order for A to stay, it needs to be greater than either B or C.

If A is less than BOTH B and C, it gets tossed.

In SAS-speak, the following doesn't appear to be correct.  Can someone please offer guidance?

if A < (B and C) then A = . ;

Using 'or' doesn't seem to work either.

Oh, and, these are numeric variables, numbers....

Hey, it's been a long day.

[KaueMAlmeida]

I haven't tried this code, but it something like that...

If A < B or A < C then A = .;

If A gt B or A gt C then A = .;

Those variables are numerical? What kind of values do you have on each column? May you will have some problem when calculating the values.

[andreas_lds]

In order for A to stay, it needs to be greater than either B or C.

Must be translated to the logic expression A > B or A > C.

If A is less than BOTH B and C

Means: A < B and A < C

Next step: what should happen if A = B and A = C?

And: why should A be set to missing if it should "stay"?

[Kurt_Bremser · Accepted solution]

if a > min(b,c);

[NKormanik]

Terrific, @Kurt_Bremser .  How easy that was.  Thanks!

Shows the great usefulness of max(), min().  And also, not().

[ballardw]

---

@NKormanik wrote:

Given A, B, C

In order for A to stay, it needs to be greater than either B or C.

If A is less than BOTH B and C, it gets tossed.

In SAS-speak, the following doesn't appear to be correct.  Can someone please offer guidance?

if A < (B and C) then A = . ;

Using 'or' doesn't seem to work either.

Oh, and, these are numeric variables, numbers....

Hey, it's been a long day.

---

It is "correct" in the way that SAS works. When you place a comparison in () such as (B and C) it evaluates that expression. Then returns the result of that expression as 1 (for true) or 0 for false. The comparison (B and C), since you use a Logic expression is testing "is B true and is C true". Which returns 1 when both are "true" or 0 if any are not "true". Values of 0 or missing are not "true".

Example:

data example;
   input b c;
   result = (b and c);
datalines;
1 1
0 1
. 1
. .
2 3
-4 5
6 7
1000000 .000004
;

So your code was asking, essentially, is "A <0" or is "A<1" depending on the values of B and C.