moe.easy_interface package

Submodules

moe.easy_interface.bandit_simple_endpoint module

Simple functions for hitting the REST endpoints of a MOE bandit service.

See Multi-Armed Bandits for more details on multi-armed bandits.

moe.easy_interface.bandit_simple_endpoint.bandit(historical_info, type='bandit_bla', rest_host='127.0.0.1', rest_port=6543, testapp=None, **kwargs)[source]

Hit the rest endpoint for allocating arms of a bandit given arms already sampled (historical info).

moe.easy_interface.experiment module

Classes for MOE optimizable experiments.

class moe.easy_interface.experiment.Experiment(domain_bounds, points_sampled=None)[source]

Bases: object

A class for MOE optimizable experiments.

build_json_payload()[source]

Construct a json serializeable and MOE REST recognizeable dictionary of the experiment.

moe.easy_interface.simple_endpoint module

Simple functions for hitting the REST endpoints of a MOE service.

moe.easy_interface.simple_endpoint.call_endpoint_with_payload(rest_host, rest_port, endpoint, json_payload, testapp=None)[source]

Send a POST request to a url with a given json_payload, return the response as a dict.

moe.easy_interface.simple_endpoint.gp_hyper_opt(points_sampled, rest_host='127.0.0.1', rest_port=6543, testapp=None, **kwargs)[source]

Hit the rest endpoint for optimizing the hyperparameters of a gaussian process, given points already sampled.

moe.easy_interface.simple_endpoint.gp_mean_var(points_sampled, points_to_evaluate, rest_host='127.0.0.1', rest_port=6543, testapp=None, **kwargs)[source]

Hit the rest endpoint for calculating the posterior mean and variance of a gaussian process, given points already sampled.

moe.easy_interface.simple_endpoint.gp_next_points(moe_experiment, method_route_name='gp_next_points_epi', rest_host='127.0.0.1', rest_port=6543, testapp=None, **kwargs)[source]

Hit the rest endpoint for finding next point of highest EI at rest_host:rest_port corresponding to the method with the given experiment.

Module contents

An easy, simple interface for using the MOE REST service.

Contains:

  • bandit_simple_endpoint.py - An example use of the bandit REST interface
  • experiment.py - The a class for historic information required to optimize using MOE
  • simple_endpoint.py - An example use of the REST interface using python