"""Test functions of solver.py"""
import numpy as np
import pytest
from cobra.exceptions import OptimizationError
from cobra.util import solver as su
[docs]stable_optlang = ["glpk", "cplex", "gurobi"]
[docs]optlang_solvers = ["optlang-" + s for s in stable_optlang if s in su.solvers]
[docs]def test_solver_list():
assert len(su.solvers) >= 1
assert "glpk" in su.solvers
[docs]def test_interface_str():
assert su.interface_to_str("nonsense") == "nonsense"
assert su.interface_to_str("optlang.glpk_interface") == "glpk"
assert su.interface_to_str("optlang-cplex") == "cplex"
[docs]def test_solver_name():
assert su.get_solver_name() == "glpk"
[docs]def test_choose_solver(model):
so = su.choose_solver(model, "glpk")
assert su.interface_to_str(so) == "glpk"
if any(s in su.solvers for s in su.qp_solvers):
qp_choice = su.choose_solver(model, qp=True)
assert su.interface_to_str(qp_choice) in su.qp_solvers
else:
with pytest.raises(su.SolverNotFound):
su.choose_solver(model, qp=True)
[docs]def test_linear_reaction_coefficients(model):
coefficients = su.linear_reaction_coefficients(model)
assert coefficients == {model.reactions.Biomass_Ecoli_core: 1}
@pytest.mark.parametrize("solver", optlang_solvers)
[docs]def test_fail_non_linear_reaction_coefficients(model, solver):
model.solver = solver
try:
model.objective = model.problem.Objective(
model.reactions.ATPM.flux_expression ** 2
)
except ValueError:
pass
else:
coefficients = su.linear_reaction_coefficients(model)
assert coefficients == {}
with pytest.raises(ValueError):
model.reactions.ACALD.objective_coefficient = 1
[docs]def test_add_remove(model):
v = model.variables
new_var = model.problem.Variable("test_var", lb=-10, ub=-10)
new_constraint = model.problem.Constraint(
v.PGK - new_var, name="test_constraint", lb=0
)
su.add_cons_vars_to_problem(model, [new_var, new_constraint])
assert "test_var" in model.variables.keys()
assert "test_constraint" in model.constraints.keys()
su.remove_cons_vars_from_problem(model, [new_var, new_constraint])
assert "test_var" not in model.variables.keys()
assert "test_constraint" not in model.constraints.keys()
[docs]def test_add_remove_in_context(model):
v = model.variables
new_var = model.problem.Variable("test_var", lb=-10, ub=-10)
with model:
su.add_cons_vars_to_problem(model, [new_var])
su.remove_cons_vars_from_problem(model, [v.PGM])
assert "test_var" in model.variables.keys()
assert "PGM" not in model.variables.keys()
assert "test_var" not in model.variables.keys()
assert "PGM" in model.variables.keys()
[docs]def test_absolute_expression(model):
v = model.variables
with model:
parts = su.add_absolute_expression(model, 2 * v.PGM, name="test", ub=100)
assert len(parts) == 3
assert "test" in model.variables.keys()
assert "abs_pos_test" in model.constraints.keys()
assert "abs_neg_test" in model.constraints.keys()
assert "test" not in model.variables.keys()
assert "abs_pos_test" not in model.constraints.keys()
assert "abs_neg_test" not in model.constraints.keys()
@pytest.mark.parametrize("solver", optlang_solvers)
[docs]def test_fix_objective_as_constraint(solver, model):
model.solver = solver
with model as m:
su.fix_objective_as_constraint(model, 1.0)
constraint_name = m.constraints[-1]
assert abs(m.constraints[-1].expression - m.objective.expression) < 1e-6
assert constraint_name not in m.constraints
su.fix_objective_as_constraint(model)
constraint_name = model.constraints[-1]
assert abs(model.constraints[-1].expression - model.objective.expression) < 1e-6
assert constraint_name in model.constraints
@pytest.mark.parametrize("solver", optlang_solvers)
[docs]def test_fix_objective_as_constraint_minimize(model, solver):
model.solver = solver
model.reactions.Biomass_Ecoli_core.bounds = (0.1, 0.1)
minimize_glucose = model.problem.Objective(
model.reactions.EX_glc__D_e.flux_expression, direction="min"
)
su.set_objective(model, minimize_glucose)
su.fix_objective_as_constraint(model)
fx_name = "fixed_objective_{}".format(model.objective.name)
constr = model.constraints
# Ensure that a solution exists on non-GLPK solvers.
model.slim_optimize()
assert (constr[fx_name].lb, constr[fx_name].ub) == (
None,
model.solver.objective.value,
)
@pytest.mark.parametrize("solver", optlang_solvers)
[docs]def test_add_lp_feasibility(model, solver):
model.solver = solver
with model:
with model:
su.add_lp_feasibility(model)
assert "s_plus_succoa_c" in model.variables
assert np.isclose(model.slim_optimize(), 0.0)
model.reactions.EX_glc__D_e.lower_bound = 1
assert np.isnan(model.slim_optimize())
assert "s_plus_succoa_c" not in model.variables
su.add_lp_feasibility(model)
assert np.isclose(model.slim_optimize(), 1.3872307692307695)
@pytest.mark.parametrize("solver", optlang_solvers)
[docs]def test_add_lexicographic_constraints(model, solver):
model.solver = solver
rxns = ["Biomass_Ecoli_core", "EX_glc__D_e", "EX_o2_e"]
with model:
out = su.add_lexicographic_constraints(model, rxns, ["max", "min", "max"])
print(model.reactions.Biomass_Ecoli_core.bounds)
assert np.isclose(model.constraints[-3].lb, out[rxns[0]])
assert np.isclose(model.constraints[-2].ub, out[rxns[1]])
assert np.isclose(model.constraints[-1].lb, out[rxns[2]])
with model:
su.add_lexicographic_constraints(model, rxns, "max")
with model:
su.add_lexicographic_constraints(model, rxns)
[docs]def test_time_limit(large_model):
if su.interface_to_str(large_model.problem) != "glpk":
pytest.skip("requires GLPK")
# It is done like this since optlang accepts inputs in seconds
# whereas GLPK accepts milliseconds
large_model.solver.configuration._smcp.tm_lim = 1
with pytest.warns(UserWarning):
sol = large_model.optimize()
assert sol.fluxes is not None
with pytest.raises(OptimizationError):
sol = large_model.optimize(raise_error=True)