if rank == 0:
"""
executed by main MPI process
mpiexec -n <num_nodes> python -m mpi4py.futures mpi\mpi_bagging.py
will create 1 dispatcher node with rank 0 and num_node-1 workers for the pool
"""
## generate Training set, Test set & V_0s
X_train, y_train = generate_train_set(Config.N_train, Config.Delta,
Config.d)
X_test, y_test = generate_test_set(Config.N_test, Config.Delta, Config.d)
V_0_train = generate_V_0(Config.N_train, Config.Delta, Config.d)
V_0_test = generate_V_0(Config.N_test, Config.Delta, Config.d)
logger.info(f"V_0_test = {V_0_test}")
reference = create_GPR(Config.N_train)
reference.fit(X_train, y_train)
f_X = reference.predict(X_test)
reference_error = normalized_error_VT(f_X, y_test, V_0_test)
logger.info(f"reference error : {reference_error}")
## MPI execute
results = []
with MPIPoolExecutor() as executor:
futures = []
# combine the ensemble predictions into a single prediction for each trial of each hyperparam setting
#
predictions = []
for result in results:
predictor_results = result[2] #M* [(m_ij,s_ij)] each of len trials
if SOFTVOTING:
bagging_predictions = trials_soft_prediction(
predictor_results, Config.trials)
else:
bagging_predictions = trials_hard_prediction(
predictor_results, Config.trials)
predictions.append([result[0], result[1], bagging_predictions])
## compute the normalized error for all hyperparam runs & all trials
V_0 = generate_V_0(100000, Config.Delta, Config.d)
logger.info(f"V_0 = {V_0}")
normalized_error_results = []
for result in predictions:
errors = []
for index, prediction in enumerate(result[2]):
error = normalized_error_VT(prediction,
DataContainer.y_test_list[index], V_0)
errors.append(error)
normalized_error_results.append([result[0], result[1], errors])
print(normalized_error_results)
if WRITEBACK:
write_results("mpi_regularized_bagging", normalized_error_results,
Config)