logging.info(
'================================================================================'
)
model = LambdaMART(
metric='NDCG@10',
max_leaf_nodes=7,
shrinkage=0.1,
estopping=30,
n_jobs=-1,
random_state=42,
use_pines=True,
pines_kwargs=dict(
switch_criterion=ObliviousCartSwitchCriterionType.OBLIVIOUS_WHILE_CAN,
tree_type=TreeType.OBLIVIOUS_CART,
max_n_splits=10,
min_samples_leaf=50,
max_depth=10,
))
model.fit(training_queries, validation_queries=validation_queries)
logging.info(
'================================================================================'
)
logging.info('%s on the test queries: %.8f' %
(model.metric, model.evaluate(test_queries, n_jobs=-1)))
model.save('LambdaMART_L7_S0.1_E50_' + model.metric)
dcg_folds_scores.append(dcg_score)
map_folds_scores.append(map_score)
#evaluate nDCG10 on test queries for best model
logging.info('%s on the test queries: %.8f' % ('NDCG@10', dcg_score))
#evaluate MAP on test queries
logging.info('%s on the test queries: %.8f' % ('MAP', map_score))
with open('folds_results.txt', 'a') as f:
f.write("Fold %d \n" % (i))
f.write("nDCG@10 %f\n" % (dcg_score))
f.write("MAP %f\n" % (map_score))
#save the model
model.save('models/LambdaMART_Fold_' + str(i))
# filename = 'models/LambdaMART_L7_S0.1_E50_nDCG@10'
# model = LambdaMART.load(filepath=filename)
#average the ndcgs and the maps
dcg_mean = np.mean(dcg_folds_scores)
#evaluate nDCG10 on test queries for best model
logging.info('Average %s on the test queries from all folds: %.8f' %
('NDCG@10', dcg_mean))
map_mean = np.mean(map_folds_scores)
#evaluate MAP on test queries
logging.info('Average %s on the test queries from all folds: %.8f' %
('MAP', map_mean))
with open('folds_results.txt', 'a') as f:
# Find constant query-document features.
cfs = find_constant_features([training_queries,
validation_queries,
test_queries])
# Get rid of constant features and (possibly) remove useless queries.
training_queries.adjust(remove_features=cfs, purge=remove_useless_queries)
validation_queries.adjust(remove_features=cfs, purge=remove_useless_queries)
test_queries.adjust(remove_features=cfs)
# Print basic info about query datasets.
logging.info('Train queries: %s' % training_queries)
logging.info('Valid queries: %s' % validation_queries)
logging.info('Test queries: %s' % test_queries)
logging.info('=' * 80)
model = LambdaMART(metric='NDCG@10', max_leaf_nodes=7, shrinkage=0.1,
estopping=50, n_jobs=-1, min_samples_leaf=50,
random_state=42)
model.fit(training_queries, validation_queries=validation_queries)
logging.info('=' * 80)
logging.info('%s on the test queries: %.8f'
% (model.metric, model.evaluate(test_queries, n_jobs=-1)))
model.save('LambdaMART_L7_S0.1_E50_' + model.metric)
