Choosing a solver¶

How to solve on a solver other than HiGHS, pass it options, or write the model to a file for a solver specsolve does not run. HiGHS ships with specsolve and is the default. Nothing in the spec names a solver, so the call chooses one.

See which solvers this build has¶

Name a solver that does not exist, and the refusal lists the ones that do. This block runs when the site is built, so the list is this commit's:

import specsolve as sps

try:
    sps.solve('examples/dispatch.yaml', {}, solver_name='cplex')
except sps.SpecsolveError as exc:
    print(exc)

unknown solver 'cplex' — this build solves with gurobi, highs, xpress.

The name is checked before any data is attached.

Install the solver¶

Every solver except HiGHS comes with the extra of the same name:

pip install "specsolve[gurobi]"

A name the build knows, but whose package is not installed, raises ModuleNotFoundError with the pip install line to run.

Name it in the call¶

solver_name= goes on solve, Model.solve and solve_over:

result = sps.solve('dispatch.yaml', sources, solver_name='gurobi')

Pass solver options¶

solver_options= goes to the solver verbatim, in its own vocabulary. A time limit is time_limit on HiGHS and TimeLimit on Gurobi:

result = sps.solve('dispatch.yaml', sources, solver_name='gurobi', solver_options={'TimeLimit': 60})

A solver that cannot take the model¶

Solvers take different constructs: HiGHS has no SOS sets, for example. solve refuses a built model the solver cannot take, before the load, and the refusal names the solvers that do take the construct. To ask without solving, build and check:

import specsolve as sps

model = sps.build('dispatch.yaml', sources)
model.check('gurobi')

The full table is what each sink takes.

The answer is read off the model the build produced, not off the file. A file may declare a quadratic cost that the data prices at zero everywhere, or an integer variable that a where: leaves without a column; that model is an LP, and every solver takes it.

Write the model to a file¶

For a solver or tool that specsolve does not run, write the model. The suffix picks the format, and this build writes .lp and .mps:

sps.write('dispatch.yaml', sources, 'dispatch.lp')

The two formats carry different constructs: MPS has no quadratic objective or indicator, for example. The table above covers both.

A written column is x0 and a row is c0 unless you ask for names. With names=True, each column and row is named by its declaration and coordinate, such as p(0,wind) and balance(0). Use it when the tool that reads the file finds a variable by name, as JuMP's variable_by_name does:

sps.write('dispatch.yaml', sources, 'dispatch.lp', names=True)

A label keeps its letters, digits and !"#$%&'.;?@`{|}~. Every other character is written as _, because an LP reader refuses it, so north sea becomes north_sea. A declaration with no dims is written total(). When two labels of one declaration become the same name, the write is refused and the error names both labels. model.check('.lp', names=True) gives the same refusal without writing a file.

Hand the model to pyomo¶

To extend the model in pyomo, or solve it with a solver pyomo runs, export the built model. It needs the [pyomo] extra:

import pyomo.environ as pyo

with sps.build('dispatch.yaml', sources) as model:
    m = model.to_pyomo()
pyo.SolverFactory('appsi_highs').solve(m)
m.p[0, 'wind'].value

Each variable is a Var and each constraint a Constraint, named as the spec declares them and indexed by their labels. A coordinate that a where removed has no entry. A row holds the numbers the build computed, so new data means a new export.

A pyomo model has one namespace. A spec may give a variable and a constraint the same name, and some names are already taken: dual, rc and slack, where pyomo's solvers write their results, the attributes every ConcreteModel has, such as write, and objective. The export does not rename anything by itself. It refuses each clash in one error, which suggests a rename to pass back:

m = model.to_pyomo(rename={'constraints': {'start_up': 'start_up_constraint'}})

model.check('pyomo', rename=...) gives the same answer without the export.