Attached is a model which ends after quite some time with the following
end lines:
+ 29790: mip = 1.300910695e-04 >= 9.430119284e-05 27.5% (253; 16770)
+ 29845: mip = 1.300910695e-04 >= 9.430119284e-05 27.5% (253; 16809)
+ 29896: mip = 1.300910695e-04 >= 9.430119284e-05 27.5% (243; 16864)
Assertion failed: x >= lb; file glpmip2.c; line 230
The execution is started with:
/usr/local/bin/glpsol --math piecewiselinearinteger.mod --intopt -o sintest.txt
As it uses the new integer optimization features I submit this report in
the hope that it might help debugging it further.
The model has worked properly for other input values (the three first
parameters *bits) earlier.
It is verified by running the same model twice and exectly the same thing
happens.
With best regards
Oscar Gustafsson
# Piecewise linear approximation
# Oscar Gustafsson, [EMAIL PROTECTED]
param inputbits := 12;
param partitionbits := 5;
param coeffbits := 12;
set Y := {0..(2^partitionbits-1)};
set Z := {0..(2^(inputbits-partitionbits)-1)};
param X{y in Y, z in Z} := (2^(inputbits-partitionbits)*y + z)/2^inputbits;
var p{y in Y} integer;
var q{y in Y} integer;
var error>=0;
s.t. postive{y in Y, z in Z}: (p[y] + z*q[y]/2^(inputbits))/2^coeffbits -
cos(X[y,z]) <= error;
s.t. negative{y in Y, z in Z}: -((p[y] + z*q[y]/2^(inputbits))/2^coeffbits -
cos(X[y,z])) <= error;
minimize e: error;
solve;
printf "p = [" ;
for {y in Y}
{
printf " %d", p[y] ;
}
printf("];\n");
printf("q = [");
for {y in Y}
{
printf "%d ", q[y];
}
printf("];\n");
end;
_______________________________________________
Bug-glpk mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-glpk
