def dispatch(self, method_name, *args, **kwargs): args_names = method_name_to_args[method_name] args_dict = dict(zip(args_names, args)) kwargs.update(args_dict) out = au.call(method_name, kwargs, self.URI) if isinstance(out, tuple): out, id = out return out
def call_bayesdb_engine(self, method_name, args_dict): if self.online: out, id = au.call(method_name, args_dict, self.URI) else: method = getattr(self.engine, method_name) argnames = inspect.getargspec(method)[0] args = [args_dict[argname] for argname in argnames if argname in args_dict] out = method(*args) return out
T, M_r, M_c = du.gen_factorial_data_objects( seed, num_clusters, num_cols, num_rows, num_splits, max_mean=max_mean, max_std=max_std ) # non-stub functions non_stub = set( ["initialize", "initialize_and_analyze", "analyze", "impute", "impute_and_confidence", "simple_predictive_sample"] ) method_name = "initialize" args_dict = dict() args_dict["M_c"] = M_c args_dict["M_r"] = M_r args_dict["T"] = T out, id = au.call(method_name, args_dict, URI) M_c, M_r, X_L_prime, X_D_prime = out method_name = "analyze" args_dict = dict() args_dict["M_c"] = M_c args_dict["T"] = T args_dict["X_L"] = X_L_prime args_dict["X_D"] = X_D_prime args_dict["kernel_list"] = () args_dict["n_steps"] = 10 args_dict["c"] = () args_dict["r"] = () args_dict["max_iterations"] = "max_iterations" args_dict["max_time"] = "max_time" out, id = au.call(method_name, args_dict, URI)
URI = 'http://' + hostname + ':8007' method_name = 'simple_predictive_sample' # views_samples = [] for replicating_samples_params in views_replicating_samples_params: this_view_samples = [] for replicating_sample_params in replicating_samples_params: args_dict = dict( M_c=save_dict['M_c'], X_L=save_dict['X_L'], X_D=save_dict['X_D'], Y=replicating_sample_params['Y'], Q=replicating_sample_params['Q'], n=replicating_sample_params['n'], ) this_view_this_sample, id = au.call(method_name, args_dict, URI) print(id) this_view_samples.extend(this_view_this_sample) views_samples.append(this_view_samples) for view_idx, view_samples in enumerate(views_samples): data_array = numpy.array(view_samples) pu.plot_T(data_array) pylab.title('simple_predictive_sample observed, view %s on remote' % view_idx) # test simple_predictive_sample_unobserved observed_Q = views_replicating_samples_params[0][0]['Q'] Q = [(int(1E6), old_tuple[1]) for old_tuple in observed_Q] new_row_samples = [] new_row_sample = su.simple_predictive_sample(M_c, X_L,
URI = 'http://' + hostname + ':8007' method_name = 'simple_predictive_sample' # views_samples = [] for replicating_samples_params in views_replicating_samples_params: this_view_samples = [] for replicating_sample_params in replicating_samples_params: args_dict = dict( M_c=save_dict['M_c'], X_L=save_dict['X_L'], X_D=save_dict['X_D'], Y=replicating_sample_params['Y'], Q=replicating_sample_params['Q'], n=replicating_sample_params['n'], ) this_view_this_sample, id = au.call( method_name, args_dict, URI) print id this_view_samples.extend(this_view_this_sample) views_samples.append(this_view_samples) for view_idx, view_samples in enumerate(views_samples): data_array = numpy.array(view_samples) pu.plot_T(data_array) pylab.title('simple_predictive_sample observed, view %s on remote' % view_idx) # test simple_predictive_sample_unobserved observed_Q = views_replicating_samples_params[0][0]['Q'] Q = [(int(1E6), old_tuple[1]) for old_tuple in observed_Q] new_row_samples = [] new_row_sample = su.simple_predictive_sample( M_c, X_L, X_D, Y, Q, get_next_seed, n=1000) new_row_samples.extend(new_row_sample)
parser = argparse.ArgumentParser() parser.add_argument('pkl_name', type=str) parser.add_argument('--hostname', default='localhost', type=str) parser.add_argument('--seed', default=0, type=int) parser.add_argument('--start_id', default=0, type=int) args = parser.parse_args() pkl_name = args.pkl_name hostname = args.hostname seed = args.seed id = args.start_id # settings URI = 'http://' + hostname + ':8007' print('URI: ', URI) save_dict = fu.unpickle(pkl_name) method_name = 'analyze' args_dict = dict() args_dict['M_c'] = save_dict['M_c'] args_dict['T'] = save_dict['T'] args_dict['X_L'] = save_dict['X_L'] args_dict['X_D'] = save_dict['X_D'] args_dict['kernel_list'] = 'kernel_list' args_dict['n_steps'] = 1 args_dict['c'] = 'c' args_dict['r'] = 'r' args_dict['max_iterations'] = 'max_iterations' args_dict['max_time'] = 'max_time' out, id = au.call(method_name, args_dict, URI, id) X_L_prime, X_D_prime = out