Parse output from a solver and updates problem state
Updates problem status, problem value, and primal and dual variable values
unpack_results(object, solution, chain, inverse_data)
object |
A Problem object. |
solution |
A Solution object. |
chain |
The corresponding solving Chain. |
inverse_data |
A InverseData object or list containing data necessary for the inversion. |
A list containing the solution to the problem:
statusThe status of the solution. Can be "optimal", "optimal_inaccurate", "infeasible", "infeasible_inaccurate", "unbounded", "unbounded_inaccurate", or "solver_error".
valueThe optimal value of the objective function.
solverThe name of the solver.
solve_timeThe time (in seconds) it took for the solver to solve the problem.
setup_timeThe time (in seconds) it took for the solver to set up the problem.
num_itersThe number of iterations the solver had to go through to find a solution.
getValueA function that takes a Variable object and retrieves its primal value.
getDualValueA function that takes a Constraint object and retrieves its dual value(s).
## Not run:
x <- Variable(2)
obj <- Minimize(x[1] + cvxr_norm(x, 1))
constraints <- list(x >= 2)
prob1 <- Problem(obj, constraints)
# Solve with ECOS.
ecos_data <- get_problem_data(prob1, "ECOS")
# Call ECOS solver interface directly
ecos_output <- ECOSolveR::ECOS_csolve(
c = ecos_data[["c"]],
G = ecos_data[["G"]],
h = ecos_data[["h"]],
dims = ecos_data[["dims"]],
A = ecos_data[["A"]],
b = ecos_data[["b"]]
)
# Unpack raw solver output.
res1 <- unpack_results(prob1, "ECOS", ecos_output)
# Without DCP validation (so be sure of your math), above is equivalent to:
# res1 <- solve(prob1, solver = "ECOS")
X <- Variable(2,2, PSD = TRUE)
Fmat <- rbind(c(1,0), c(0,-1))
obj <- Minimize(sum_squares(X - Fmat))
prob2 <- Problem(obj)
scs_data <- get_problem_data(prob2, "SCS")
scs_output <- scs::scs(
A = scs_data[['A']],
b = scs_data[['b']],
obj = scs_data[['c']],
cone = scs_data[['dims']]
)
res2 <- unpack_results(prob2, "SCS", scs_output)
# Without DCP validation (so be sure of your math), above is equivalent to:
# res2 <- solve(prob2, solver = "SCS")
## End(Not run)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.