Scalinea

Summary

This project provides a mathematical programming modeling library. It aims to provide a DSL for the creation of linear and quadratic optimization model. Solution will be solve with CBC, lp_solve, glpk and Gurobi.

Current version

• Basic DSL to create a model
• Solution provided by CBC-Coin Or and Gurobi

This current version is usable but is subject to change

Example

val a = IVar("a")
val b = IVar("b")
val c = IVar("c")

val system = {
  dsl.System.define.constraints(
    500*a + 1200*b + 1500*c <= 10000,
    a <= b
  ).maximize(
    10.0*a + 20.0*b
  ).build
}

val solver: Solver = CbcLpSolver // or GurobiSolver
solver.solve(system) match {
  case Success(sol: Solution, _) => {
    println("Optimal: " + sol.isOptimal )
    for( v <- Seq(a,b,c) ) {
      println( s"${v.symbol}: ${sol(v)}" )
    }
  }
  case _ => println("oups")
}

News

DSL for binary variables is now possible

  val system = {
    dsl.System.define.constraints(
      x `imply` z,
      z `iif` w,
      z & w,
      x | y `imply` exactlyOneOf(z1, z2 | z3, z1 & z2, z4 `iif` z5)
      ...

    ).maximize(
      ...
    ).build

Next step

• add new solvers (glpk, lp_solve, Cplex)
• improve existing solver solution parsing with explicit status (Unboundable, Unfeasible, ...)
• add more tests and more examples
• create documentation
• publish the first beta release :)