I have a model with about 2000 equations. If I try to solve the model with the commands:
Code: Select all
{%model}.solve(o=n)
freeze(_a1) {%model}.msg
{%model}.solve(o=g)
freeze(_a2) {%model}.msg
{%model}.solve(o=b)
freeze(_a3) {%model}.msg
%sp="_results1"
spool {%sp}
{%sp}.append _a1
{%sp}.append _a2
{%sp}.append _a3
the model will not solve. However if I alter the commands to:
Code: Select all
{%model}.solve(o=n,v=f, d=d)
freeze(_a4) {%model}.msg
{%model}.solve(o=g,v=f, d=d)
freeze(_a5) {%model}.msg
{%model}.solve(o=b,v=f, d=d)
freeze(_a6) {%model}.msg
%sp="_results2"
spool {%sp}
{%sp}.append _a4
{%sp}.append _a5
{%sp}.append _a6
the model solves. (see the two attachments).
I find this very strange. Why is this?
ideally, a user should just have to use a "solve" command and the program should by itself try different solution algorithms (and options) to find the one that solves the model. Is this something you can implement?
Thomas