LOQO Options#
ampl: option solver loqo; # change the solver
ampl: option loqo_options 'option1=value1 option2=value2'; # specify options
ampl: solve; # solve the problem
Solver options obtained with $ loqo -=
.
bndpush initial distance from bounds
bounds name of MPS BOUNDS section
convex assert problem is convex
dense dense window threshold
dual dual-favored matrix reordering
epsdiag eps tol diag perturbation
epsnum eps tol for numerical factorization
epssol eps tol for forward/backward solve
honor_bnds honor bounds on variables
honor_bnds_init honor bnds initially
ignore_initsol ignore given initial guesses; make automatic choice
inftol Feasibility tolerance
inftol2 Feasibility tolerance 2
iterlim Iteration Limit (ITERLIM)
lp_only linearize only---don't quadraticize
max maximize the objective
maximize maximize the objective
maxit Iteration Limit (ITERLIM)
min minimize the objective
mindeg minimum degree matrix reordering
minimize minimize the objective
minlocfil minimum local fill matrix reordering
mufactor mu factor
noreord no matrix reordering
obj name of MPS objective row
objno objective number: 1= first, 0= none
outlev output level
penalty penalty parameter
pred_corr pred= 0, pred_corr= 1, unset= -1
primal primal-favored matrix reordering
ranges name of MPS RANGES section
rhs name of MPS right-hand side section
sdp assert problem is an SDP
sigfig significant digits (agreement in primal and dual objectives)
stablty mixing factor for matrix reordering
steplen step length reduction factor
timing timing destination: 1 = stdout, 2 = stderr, 3 = both
timlim Time limit
verbose synonym for outlev
version single-word phrase: report version before solving
wantsol solution report without -AMPL: sum of
1 ==> write .sol file
2 ==> print primal variable values
4 ==> print dual variable values
8 ==> do not print solution message
zero_initsol start at the origin