In SAS, we have the ability to perform and apply bitwise logical operations. There are numerous bitwise logical functions available within SAS, and I would like to highlight a few of them for you.
To understand these functions, it is necessary to understand logical AND, OR and XOR operations.
· Logical AND : Returns true if both operands are true.
· Logical OR : Returns true if at
least one of the operands is true.
· Logical XOR : Returns true if only one of the operands is
true.
SAS
offers build in functions that allow us to apply these logical operations.
These are:
·
BAND
·
BOR
·
BXOR
Code to generate bitwise logical table:
data Input;
input x y;
datalines;
0 0
0 1
1 0
1 1
;
run;
data table;
set Input;
x_and_y = BAND(x,y);
x_or_y = BOR(x,y);
x_xor_y = BXOR(x,y);
run;
proc print noobs; run;
Output:
|
x |
y |
x and y |
x or y |
x xor y |
|
0 |
0 |
0 |
0 |
0 |
|
0 |
1 |
0 |
1 |
1 |
|
1 |
0 |
0 |
1 |
1 |
|
1 |
1 |
1 |
1 |
0 |
Now
if you are wondering where we can use these functions, then do not worry I will
explain to you a few cases where these functions are super useful.
Dataset:
Consider
the below patient medication dataset contains information on the medications
given to five patients. Each row in the dataset represents the medication
status for medication A and medication B.
·
USUBJID: A unique
identifier for the patient.
·
medicine_A, Medicine_B:
A binary variable that indicates whether the patient is taking medicine A, B. 1
indicates the patient is taking medicine and 0 indicates the patient is not
taking medicine.
|
USUBJID |
medicine_A |
medicine_B |
|
1001 |
1 |
0 |
|
1002 |
0 |
1 |
|
1003 |
1 |
1 |
|
1004 |
0 |
0 |
|
1005 |
1 |
0 |
Dataset code:
data Medication;
length USUBJID $10 medicine_A $1 medicine_B $1;
informat USUBJID $10.;
input USUBJID medicine_A medicine_B;
datalines;
1001 1 0
1002 0 1
1003 1 1
1004 0 0
1005 1 0
;
run;
Questions:
1.
If I
will ask you to create a ABflag variable that flags patients who are taking
either medication.
Solution: We can accomplish this using BOR function
data ABflag;
set
medication;
ABflag = BOR(medicine_A, medicine_B);
run;
proc print noobs; run;
|
USUBJID |
medicine_A |
medicine_B |
ABflag |
|
1001 |
1 |
0 |
1 |
|
1002 |
0 |
1 |
1 |
|
1003 |
1 |
1 |
1 |
|
1004 |
0 |
0 |
0 |
|
1005 |
1 |
0 |
1 |
2.
To create
a flag variable that shows if a patient has taken both medications.
Solution: We can use the BAND function here.
data ABflag;
set
medication;
ABflag = BAND(medicine_A, medicine_B);
run;
proc print noobs; run;
|
USUBJID |
medicine_A |
medicine_B |
ABflag |
|
1001 |
1 |
0 |
0 |
|
1002 |
0 |
1 |
0 |
|
1003 |
1 |
1 |
1 |
|
1004 |
0 |
0 |
0 |
|
1005 |
1 |
0 |
0 |
3.
Similarly,
to create a flag variable that indicates if a patient has either taken or not
taken either medication A or medication B.
Solution: We can use the BXOR function.
data ABflag;
set
medication;
ABflag = BXOR(medicine_A, medicine_B);
run;
proc print noobs; run;
|
USUBJID |
medicine_A |
medicine_B |
ABflag |
|
1001 |
1 |
0 |
1 |
|
1002 |
0 |
1 |
1 |
|
1003 |
1 |
1 |
0 |
|
1004 |
0 |
0 |
0 |
|
1005 |
1 |
0 |
1 |
