/* linear fractional programming problem (NLP) */
proc optmodel;
var x {1..2} >= 0;
max z = (x[1] + 3) / (x[2] + 1);
con c1: -x[1] + x[2] <= 1;
con c2: 2*x[1] <= 3;
con c3: 3*x[1] + 2*x[2] <= 7;
solve;
print x;
quit;
/* linear programming transformation from Das/Mandal paper (LP) */
proc optmodel;
var y {1..2} >= 0;
max z = y[1] - y[2] + 3;
con c1: -y[1] + 2*y[2] <= 1;
con c2: 2*y[1] <= 3;
con c3: 3*y[1] + 9*y[2] <= 7;
solve;
print y;
quit;
/* linear programming problem from Charnes/Cooper transformation (LP) */
/* https://en.wikipedia.org/wiki/Linear-fractional_programming#Transformation_to_a_linear_program */
proc optmodel;
var y {1..2} >= 0;
var t >= 0;
max z = y[1] + 3*t;
con c1: -y[1] + y[2] <= t;
con c2: 2*y[1] <= 3*t;
con c3: 3*y[1] + 2*y[2] <= 7*t;
con c4: y[2] + t = 1;
solve;
print y;
impvar x {j in 1..2} = y[j].sol / t.sol;
print x;
quit;
SAS/IML also can do that . If you want IML solution, post it at IML forum. Rick is there.
April 27 – 30 | Gaylord Texan | Grapevine, Texas
Registration is open
Walk in ready to learn. Walk out ready to deliver. This is the data and AI conference you can't afford to miss. Register now and lock in 2025 pricing—just $495!