anml.solvers package

Submodules

Base Solvers

Basic solvers for optimization. Takes a model directly.

class ClosedFormSolver(model_instance=None)[source]

Bases: anml.solvers.interface.Solver

Solver using closed formed solution defined in corresponding model.

fit(x_init=None, data=None, options=None)[source]
class IPOPTSolver(model_instance=None)[source]

Bases: anml.solvers.interface.Solver

Solver using IPOPT

fit(x_init, data=None, options=None)[source]
class ScipyOpt(*args, **kwargs)[source]

Bases: anml.solvers.interface.Solver

A concrete class of Solver that use scipy optimize.

fit(x_init, data=None, options=None)[source]
class MultipleInitializations(sample_fun, solver=None)[source]

Bases: anml.solvers.interface.CompositeSolver

Solver with multiple initialization

fit(x_init=None, data=None, options=None)[source]
predict(**kwargs)[source]
class TrimmingSolver(*args, **kwargs)[source]

Bases: anml.solvers.interface.CompositeSolver

c_simplex(w, h)[source]
fit(x_init, data, n, options=None, pct_trimming=0.0, step_size=1.0, max_iter=100, tol=1e-06)[source]
predict(**kwargs)[source]

Solvers Interface

class CompositeSolver(solvers_list=None)[source]

Bases: anml.solvers.interface.Solver

add_solver(solver)[source]
assert_model_defined()[source]
assert_solvers_defined()[source]
property model
property solvers
exception ModelNotDefinedError[source]

Bases: anml.exceptions.ANMLError

class Solver(model_instance=None)[source]

Bases: object

assert_model_defined()[source]
fit(x_init=None, data=None, options=None, **kwargs)[source]
property model
predict(**kwargs)[source]
exception SolverNotDefinedError[source]

Bases: anml.exceptions.ANMLError

Module contents