LINDO API 7.0
Introduction
LINDO Systems has begun shipping a new release of LINDO API that includes new features to allow users to incorporate uncertainty into their optimization models. LINDO API 7.0 also includes enhancements to the linear, integer and global solvers.
Stochastic Solver Improvements Include:
- Extended API functionality to support Chance-Constrained Programs (CCP).
- Improved methods to induce correlations among stochastic parameters.
LP Solver Improvements Include:
- Sprint option to efficiently solve "skinny" LP's, having millions of variables.
MIP Solver Improvements Include:
- Significant improvements in heuristics for quickly finding good integer solutions.
- Improved identification of special structures, as in multi-period models, and exploit to achieve significant reductions in solve times.
Global Solver Improvements Include:
- Improved heuristics for finding a good feasible solution quickly.
- User can identify a constraint as being convex, if the global solver might not otherwise be able to identify it as convex. This can speed up proof of global optimality.
- Improved ability to identify constraints that can reformulated as second order cone (SOC) constraints and thus be solved by fast SOC solver.
Other Solvers, Math Library and Utilities
- More than two dozen new math functions to support various PDF, CDF, and Inverse-CDF of probability distributions.
- New solver based on Benders decomposition to solve large-scale LPs with relatively small number of linking columns. MILPs whose integer variables appear only in linking columns can also be tackled with this solver.
- Callable routine to identify (almost) lower triangular matrix structure.