diff --git a/TODO.md b/TODO.md
index 5ea6cc5cf2c933eed2e7ffbf2567d4fe812412cf..b94e576024a5294b6eaffaaf1af5003f0e034313 100644
--- a/TODO.md
+++ b/TODO.md
@@ -1,7 +1,3 @@
-* Fix pickle loading of ModelRawResults, because saving the model_object leads import issues.
-* Fix ModelFactory.load function.
* Fix model results loading in compute_results.py.
* Check that omp multiclasses classifier is working as expected.
-* In the bayesian search computation, output a different file name depending on the task of the trained model.
-* Check the best params scores of the regressors (neg_mean_squared_error leads to huge negative values).
-* Prepare the json experiment files to run.
\ No newline at end of file
+* Fix the dataset error of fetcher when job_number > 1.
\ No newline at end of file
diff --git a/code/bolsonaro/data/dataset_loader.py b/code/bolsonaro/data/dataset_loader.py
index 01d71e2d650aff597b7cb6edb8de6d62f6b419cb..8ffbc76b145ffddcf7e63871864471efc76c9737 100644
--- a/code/bolsonaro/data/dataset_loader.py
+++ b/code/bolsonaro/data/dataset_loader.py
@@ -19,7 +19,7 @@ class DatasetLoader(object):
DEFAULT_NORMALIZE_D = False
DEFAULT_DATASET_NORMALIZER = 'standard'
DEFAULT_FOREST_SIZE = 100
- DEFAULT_EXTRACTED_FOREST_SIZE = 10
+ DEFAULT_EXTRACTED_FOREST_SIZE_SAMPLES = 4
DEFAULT_DEV_SIZE = 0.2
DEFAULT_TEST_SIZE = 0.2
DEFAULT_RANDOM_SEED_NUMBER = 1
diff --git a/code/bolsonaro/utils.py b/code/bolsonaro/utils.py
index 9dff06ac65a726642cc5efe2e6ed8b8f78f40b29..797f3005c97099cb88ae81f59178310f6c078685 100644
--- a/code/bolsonaro/utils.py
+++ b/code/bolsonaro/utils.py
@@ -79,7 +79,6 @@ def change_binary_func_load(base_load_function):
return X, y
return func_load
-
@contextlib.contextmanager
def tqdm_joblib(tqdm_object):
"""Context manager to patch joblib to report into tqdm progress bar given as argument"""
@@ -100,3 +99,17 @@ def tqdm_joblib(tqdm_object):
finally:
joblib.parallel.BatchCompletionCallBack = old_batch_callback
tqdm_object.close()
+
+def is_int(value):
+ try:
+ int(value)
+ return True
+ except ValueError:
+ return False
+
+def is_float(value):
+ try:
+ float(value)
+ return True
+ except ValueError:
+ return False
diff --git a/code/compute_hyperparameters.py b/code/compute_hyperparameters.py
index 548a1a4e82c720b711859abd57b0ddcf295e5835..9135c00d1ec0f86256881952d67c8cc86a62fe5a 100644
--- a/code/compute_hyperparameters.py
+++ b/code/compute_hyperparameters.py
@@ -4,7 +4,7 @@ from bolsonaro.data.dataset_parameters import DatasetParameters
from bolsonaro.data.task import Task
from bolsonaro.error_handling.logger_factory import LoggerFactory
from bolsonaro.hyperparameter_searcher import HyperparameterSearcher
-from bolsonaro.utils import save_obj_to_json, tqdm_joblib
+from bolsonaro.utils import save_obj_to_json, tqdm_joblib, is_int, is_float
import argparse
import os
@@ -68,7 +68,7 @@ def process_job(dataset_name, seed, param_space, args):
def run_hyperparameter_search_jobs(seeds, dataset_name, param_space, args):
# Run one hyperparameter search job per seed
with tqdm_joblib(tqdm(total=len(seeds), disable=not args.verbose)) as progress_bar:
- opt_results = Parallel(n_jobs=-1)(delayed(process_job)(
+ opt_results = Parallel(n_jobs=args.job_number)(delayed(process_job)(
dataset_name, seeds[i], param_space, args) for i in range(len(seeds)))
return opt_results
@@ -108,6 +108,10 @@ def compute_best_params_over_seeds(seeds, dataset_name, param_space, args):
split = element.split('_')
param, value = '_'.join(split[:-1]), split[-1]
if param not in best_params:
+ if is_int(value):
+ value = int(value)
+ elif is_float(value):
+ value = float(value)
best_params[param] = value
if len(best_params) == len(all_param_names):
break
@@ -128,6 +132,7 @@ if __name__ == "__main__":
DEFAULT_CV = 3
DEFAULT_N_ITER = 50
DEFAULT_VERBOSE = False
+ DEFAULT_JOB_NUMBER = -1
DICT_PARAM_SPACE = {'n_estimators': Integer(10, 1000),
'min_samples_leaf': Integer(1, 1000),
'max_depth': Integer(1, 20),
@@ -144,6 +149,7 @@ if __name__ == "__main__":
parser.add_argument('--use_variable_seed_number', action='store_true', default=DEFAULT_USE_VARIABLE_SEED_NUMBER, help='Compute the amount of random seeds depending on the dataset.')
parser.add_argument('--datasets', nargs='+', type=str, default=DatasetLoader.dataset_names, help='Specify the dataset used by the estimator.')
parser.add_argument('--verbose', action='store_true', default=DEFAULT_VERBOSE, help='Print tqdm progress bar.')
+ parser.add_argument('--job_number', nargs='?', type=int, default=DEFAULT_JOB_NUMBER, help='Specify the number of job used during the parallelisation across seeds.')
args = parser.parse_args()
logger = LoggerFactory.create(LOG_PATH, os.path.basename(__file__))
diff --git a/code/compute_results.py b/code/compute_results.py
index 64124af70954cc6af6a923f03f5a122a75f453fb..5e65a775cac80b8b2262a69431312a278dad9340 100644
--- a/code/compute_results.py
+++ b/code/compute_results.py
@@ -16,147 +16,31 @@ if __name__ == "__main__":
DEFAULT_RESULTS_DIR = os.environ["project_dir"] + os.sep + 'results'
DEFAULT_MODELS_DIR = os.environ["project_dir"] + os.sep + 'models'
- DEFAULT_EXPERIMENT_IDS = None
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
+ parser.add_argument('--stage_number', nargs='?', type=int, required=True, help='Specify the stage number among [1, 4].')
+ parser.add_argument('--experiment_ids', nargs='+', type=int, required=True, help='Compute the results of the specified experiment id(s).')
parser.add_argument('--results_dir', nargs='?', type=str, default=DEFAULT_RESULTS_DIR, help='The output directory of the results.')
parser.add_argument('--models_dir', nargs='?', type=str, default=DEFAULT_MODELS_DIR, help='The output directory of the trained models.')
- parser.add_argument('--experiment_ids', nargs='+', type=int, default=DEFAULT_EXPERIMENT_IDS, help='Compute the results of the specified experiment id(s)')
args = parser.parse_args()
+ if int(args.stage_number) not in list(range(1, 5)):
+ raise ValueError('stage_number must be a supported stage id (i.e. [1, 4]).')
+
# Create recursively the results dir tree
pathlib.Path(args.results_dir).mkdir(parents=True, exist_ok=True)
- """
- Use specified list of experiments ids if availabe.
- Otherwise, list all existing experiment ids from
- the specified models directory.
- """
- experiments_ids = [str(experiment_id) for experiment_id in args.experiment_ids] \
- if args.experiment_ids is not None \
- else os.listdir(args.models_dir)
+
"""
- Raise an error if there's no experiments ids found both
- in parameter or in models directory.
+ TODO:
+ For each dataset:
+ Stage 1) A figure for the selection of the best base forest model hyperparameters (best vs default/random hyperparams)
+ Stage 2) A figure for the selection of the best combination of normalization: D normalization vs weights normalization (4 combinations)
+ Stage 3) A figure for the selection of the most relevant subsets combination: train,dev vs train+dev,train+dev vs train,train+dev
+ Stage 4) A figure to finally compare the perf of our approach using the previous selected
+ parameters vs the baseline vs other papers using different extracted forest size
+ (percentage of the tree size found previously in best hyperparams search) on the abscissa.
+
+ IMPORTANT: Compare experiments that used the same seeds among them (except for stage 1).
"""
- if experiments_ids is None or len(experiments_ids) == 0:
- raise ValueError("No experiment id was found or specified.")
-
- # Compute the plots for each experiment id
- for experiment_id in experiments_ids:
- experiment_id_path = args.models_dir + os.sep + experiment_id # models/{experiment_id}
- # Create recursively the tree results/{experiment_id}
- pathlib.Path(args.results_dir + os.sep + experiment_id).mkdir(parents=True, exist_ok=True)
- experiment_seed_root_path = experiment_id_path + os.sep + 'seeds' # models/{experiment_id}/seeds
-
- """
- Dictionaries to temporarly store the scalar results with the following structure:
- {seed_1: [score_1, ..., score_m], ... seed_n: [score_1, ..., score_k]}
- TODO: to complete to retreive more results
- """
- experiment_train_scores = dict()
- experiment_dev_scores = dict()
- experiment_test_scores = dict()
-
- experiment_weights = dict()
-
- # Used to check if all losses were computed using the same metric (it should be the case)
- experiment_score_metrics = list()
-
- # For each seed results stored in models/{experiment_id}/seeds
- for seed in os.listdir(experiment_seed_root_path):
- experiment_seed_path = experiment_seed_root_path + os.sep + seed # models/{experiment_id}/seeds/{seed}
- dataset_parameters = DatasetParameters.load(experiment_seed_path, experiment_id) # Load the dataset parameters of this experiment, with this specific seed
- dataset = DatasetLoader.load(dataset_parameters) # Load the dataset using the previously loaded dataset parameters
- extracted_forest_size_root_path = experiment_seed_path + os.sep + 'extracted_forest_size' # models/{experiment_id}/seeds/{seed}/extracted_forest_size
-
- # {{seed}:[]}
- experiment_train_scores[seed] = list()
- experiment_dev_scores[seed] = list()
- experiment_test_scores[seed] = list()
-
- experiment_weights[seed] = list()
-
- # List the forest sizes in models/{experiment_id}/seeds/{seed}/extracted_forest_size
- extracted_forest_sizes = os.listdir(extracted_forest_size_root_path)
- for extracted_forest_size in extracted_forest_sizes:
- # models/{experiment_id}/seeds/{seed}/extracted_forest_size/{extracted_forest_size}
- extracted_forest_size_path = extracted_forest_size_root_path + os.sep + extracted_forest_size
- # Load models/{experiment_id}/seeds/{seed}/extracted_forest_size/{extracted_forest_size}/model_raw_results.pickle file
- model_raw_results = ModelRawResults.load(extracted_forest_size_path)
- # Load [...]/model_parameters.json file and build the model using these parameters and the weights and forest from model_raw_results.pickle
- model = ModelFactory.load(dataset.task, extracted_forest_size_path, experiment_id, model_raw_results)
- # Save temporarly some raw results (TODO: to complete to retreive more results)
- # Save the scores
- experiment_train_scores[seed].append(model_raw_results.train_score)
- experiment_dev_scores[seed].append(model_raw_results.dev_score)
- experiment_test_scores[seed].append(model_raw_results.test_score)
- # Save the weights
- experiment_weights[seed].append(model_raw_results.weights)
- # Save the metric
- experiment_score_metrics.append(model_raw_results.score_metric)
-
- if len(set(experiment_score_metrics)) > 1:
- raise ValueError("The metrics used to compute the dev score aren't the same everytime")
-
- """
- Example of plot that just plots the losses computed
- on the train, dev and test subsets using a trained
- model, with the CI, and depending on the extracted
- forest size.
- """
- Plotter.plot_losses(
- file_path=args.results_dir + os.sep + experiment_id + os.sep + 'losses.png',
- all_experiment_scores=[experiment_train_scores, experiment_dev_scores, experiment_test_scores],
- x_value=extracted_forest_sizes,
- xlabel='Number of trees extracted',
- ylabel=experiment_score_metrics[0],
- all_labels=['train', 'dev', 'test'],
- title='Loss values of the trained model'
- )
-
- """
- TODO:
- For each dataset:
- Stage 1) A figure for the selection of the best base forest model hyperparameters (best vs default/random hyperparams)
- Stage 2) A figure for the selection of the best dataset normalization method
- Stage 3) A figure for the selection of the best combination of dataset: normalization vs D normalization vs weights normalization
- Stage 4) A figure for the selection of the most relevant subsets combination: train,dev vs train+dev,train+dev vs train,train+dev
- Stage 5) A figure for the selection of the best extracted forest size?
- Stage 6) A figure to finally compare the perf of our approach using the previous selected parameters vs the baseline vs other papers
-
- Stage 3)
- In all axis:
- - untrained forest
- - trained base forest (straight line cause it doesn't depend on the number of extracted trees)
-
- Axis 1:
- - test with forest on train+dev and OMP on train+dev
- - test with forest on train+dev and OMP on train+dev with dataset normalization
- - test with forest on train+dev and OMP on train+dev with dataset normalization + D normalization
- - test with forest on train+dev and OMP on train+dev with dataset normalization + weights normalization
- - test with forest on train+dev and OMP on train+dev with dataset normalization + D normalization + weights normalization
-
- Axis 2:
- - test with forest on train and OMP on dev
- - test with forest on train and OMP on dev with dataset normalization
- - test with forest on train and OMP on dev with dataset normalization + D normalization
- - test with forest on train and OMP on dev with dataset normalization + weights normalization
- - test with forest on train and OMP on dev with dataset normalization + D normalization + weights normalization
-
- Axis 3:
- - test with forest on train and OMP train+dev
- - test with forest on train and OMP train+dev with dataset normalization
- - test with forest on train and OMP train+dev with dataset normalization + D normalization
- - test with forest on train and OMP train+dev with dataset normalization + weights normalization
- - test with forest on train and OMP train+dev with dataset normalization + D normalization + weights normalization
-
- IMPORTANT: Same seeds used in all axis.
- """
-
- # Plot the density of the weights
- Plotter.weight_density(
- file_path=args.results_dir + os.sep + experiment_id + os.sep + 'density_weight.png',
- all_experiment_weights=experiment_weights
- )
diff --git a/code/train.py b/code/train.py
index 38f2887e16de5a0967500a304034b5934dcd4a5d..13216acda3a88b3718d765ac3a6f424efd28de76 100644
--- a/code/train.py
+++ b/code/train.py
@@ -17,6 +17,7 @@ from joblib import Parallel, delayed
import threading
import json
from tqdm import tqdm
+import numpy as np
def process_job(seed, parameters, experiment_id, hyperparameters):
@@ -82,6 +83,8 @@ if __name__ == "__main__":
# the models will be stored in a directory structure like: models/{experiment_id}/seeds/{seed_nb}/extracted_forest_size/{nb_extracted_trees}
DEFAULT_MODELS_DIR = os.environ['project_dir'] + os.sep + 'models'
DEFAULT_VERBOSE = False
+ DEFAULT_SKIP_BEST_HYPERPARAMS = False
+ DEFAULT_JOB_NUMBER = -1
begin_random_seed_range = 1
end_random_seed_range = 2000
@@ -92,8 +95,8 @@ if __name__ == "__main__":
parser.add_argument('--dataset_name', nargs='?', type=str, default=DatasetLoader.DEFAULT_DATASET_NAME, help='Specify the dataset. Regression: boston, diabetes, linnerud, california_housing. Classification: iris, digits, wine, breast_cancer, olivetti_faces, 20newsgroups, 20newsgroups_vectorized, lfw_people, lfw_pairs, covtype, rcv1, kddcup99.')
parser.add_argument('--normalize_D', action='store_true', default=DatasetLoader.DEFAULT_NORMALIZE_D, help='Specify if we want to normalize the prediction of the forest by doing the L2 division of the pred vectors.')
parser.add_argument('--dataset_normalizer', nargs='?', type=str, default=DatasetLoader.DEFAULT_DATASET_NORMALIZER, help='Specify which dataset normalizer use (either standard, minmax, robust or normalizer).')
- parser.add_argument('--forest_size', nargs='?', type=int, default=DatasetLoader.DEFAULT_FOREST_SIZE, help='The number of trees of the random forest.')
- parser.add_argument('--extracted_forest_size', nargs='+', type=int, default=DatasetLoader.DEFAULT_EXTRACTED_FOREST_SIZE, help='The number of trees selected by OMP.')
+ parser.add_argument('--forest_size', nargs='?', type=int, default=None, help='The number of trees of the random forest.')
+ parser.add_argument('--extracted_forest_size_samples', nargs='?', type=int, default=DatasetLoader.DEFAULT_EXTRACTED_FOREST_SIZE_SAMPLES, help='The number of extracted forest sizes (proportional to the forest size) selected by OMP.')
parser.add_argument('--models_dir', nargs='?', type=str, default=DEFAULT_MODELS_DIR, help='The output directory of the trained models.')
parser.add_argument('--dev_size', nargs='?', type=float, default=DatasetLoader.DEFAULT_DEV_SIZE, help='Dev subset ratio.')
parser.add_argument('--test_size', nargs='?', type=float, default=DatasetLoader.DEFAULT_TEST_SIZE, help='Test subset ratio.')
@@ -102,6 +105,9 @@ if __name__ == "__main__":
parser.add_argument('--subsets_used', nargs='+', type=str, default=DatasetLoader.DEFAULT_SUBSETS_USED, help='train,dev: forest on train, OMP on dev. train+dev,train+dev: both forest and OMP on train+dev. train,train+dev: forest on train+dev and OMP on dev.')
parser.add_argument('--normalize_weights', action='store_true', default=DatasetLoader.DEFAULT_NORMALIZE_WEIGHTS, help='Divide the predictions by the weights sum.')
parser.add_argument('--verbose', action='store_true', default=DEFAULT_VERBOSE, help='Print tqdm progress bar.')
+ parser.add_argument('--skip_best_hyperparams', action='store_true', default=DEFAULT_SKIP_BEST_HYPERPARAMS, help='Do not use the best hyperparameters if there exist.')
+ parser.add_argument('--save_experiment_configuration', nargs='+', default=None, help='Save the experiment parameters specified in the command line in a file. Args: {{stage_num}} {{name}}')
+ parser.add_argument('--job_number', nargs='?', type=int, default=DEFAULT_JOB_NUMBER, help='Specify the number of job used during the parallelisation across seeds.')
args = parser.parse_args()
if args.experiment_configuration:
@@ -115,22 +121,31 @@ if __name__ == "__main__":
logger = LoggerFactory.create(LOG_PATH, os.path.basename(__file__))
- # The number of tree to extract from forest (K)
- parameters['extracted_forest_size'] = parameters['extracted_forest_size'] \
- if type(parameters['extracted_forest_size']) == list \
- else [parameters['extracted_forest_size']]
-
hyperparameters_path = os.path.join('experiments', args.dataset_name, 'stage1', 'params.json')
- if os.path.exists(hyperparameters_path):
+ if os.path.exists(hyperparameters_path) and not args.skip_best_hyperparams:
logger.info("Hyperparameters found for this dataset at '{}'".format(hyperparameters_path))
with open(hyperparameters_path, 'r+') as file_hyperparameter:
hyperparameters = json.load(file_hyperparameter)['best_parameters']
else:
hyperparameters = {}
- if parameters['forest_size'] is not None:
+ """
+ First case: no best hyperparameters are specified and no forest_size parameter
+ specified in argument, so use the DEFAULT_FOREST_SIZE.
+ Second case: no matter if hyperparameters are specified, the forest_size parameter
+ will override it.
+ Third implicit case: use the number of estimators found in specified hyperparameters.
+ """
+ if len(hyperparameters) == 0 and parameters['forest_size'] is None:
+ hyperparameters['n_estimators'] = DatasetLoader.DEFAULT_FOREST_SIZE
+ elif parameters['forest_size'] is not None:
hyperparameters['n_estimators'] = parameters['forest_size']
+ # The number of tree to extract from forest (K)
+ parameters['extracted_forest_size'] = [int(hyperparameters['n_estimators'] * coeff) \
+ for coeff in np.linspace(0, 1, parameters['extracted_forest_size_samples'] + 1,
+ endpoint=False)[1:]]
+
if parameters['seeds'] != None and parameters['random_seed_number'] > 1:
logger.warning('seeds and random_seed_number parameters are both specified. Seeds will be used.')
@@ -142,15 +157,29 @@ if __name__ == "__main__":
# Resolve the next experiment id number (last id + 1)
experiment_id = resolve_experiment_id(parameters['models_dir'])
logger.info('Experiment id: {}'.format(experiment_id))
+ parameters['experiment_id'] = experiment_id
"""
If the experiment configuration isn't coming from
an already existing file, save it to a json file to
- keep trace of it.
+ keep trace of it (either a specified path, either in 'unnamed' dir.).
"""
if args.experiment_configuration is None:
- with open(args.experiment_configuration_path + os.sep + 'unnamed' + os.sep + 'unnamed_{}.json'.format(
- experiment_id), 'w') as output_file:
+ if args.save_experiment_configuration:
+ if len(args.save_experiment_configuration) != 2:
+ raise ValueError('save_experiment_configuration must have two parameters.')
+ elif int(args.save_experiment_configuration[0]) not in list(range(1, 5)):
+ raise ValueError('save_experiment_configuration first parameter must be a supported stage id (i.e. [1, 4]).')
+ output_experiment_configuration_path = os.path.join(args.experiment_configuration_path,
+ args.dataset_name, 'stage' + args.save_experiment_configuration[0],
+ args.save_experiment_configuration[1] + '_{}.json'.format(
+ experiment_id))
+ else:
+ pathlib.Path(os.path.join(args.experiment_configuration_path, 'unnamed')).mkdir(parents=True, exist_ok=True)
+ output_experiment_configuration_path = os.path.join(
+ args.experiment_configuration_path, 'unnamed', 'unnamed_{}.json'.format(
+ experiment_id))
+ with open(output_experiment_configuration_path, 'w') as output_file:
json.dump(
parameters,
output_file,
@@ -159,5 +188,5 @@ if __name__ == "__main__":
# Run as much job as there are seeds
with tqdm_joblib(tqdm(total=len(seeds), disable=not args.verbose)) as progress_bar:
- Parallel(n_jobs=-1)(delayed(process_job)(seeds[i],
+ Parallel(n_jobs=args.job_number)(delayed(process_job)(seeds[i],
parameters, experiment_id, hyperparameters) for i in range(len(seeds)))
diff --git a/experiments/boston/stage1/params.json b/experiments/boston/stage1/params.json
index f2f3abbe37d05ef6123ce819ebee4dfac2b254a7..8b530a1131567dc32b0e6274dbda45f143ce4cab 100644
--- a/experiments/boston/stage1/params.json
+++ b/experiments/boston/stage1/params.json
@@ -4,9 +4,9 @@
"best_score_test": -13.650326577972058,
"best_parameters": {
"max_features": "auto",
- "min_samples_leaf": "1",
- "max_depth": "20",
- "n_estimators": "1000"
+ "min_samples_leaf": 1,
+ "max_depth": 20,
+ "n_estimators": 1000
},
"random_seed": [
1812,
diff --git a/experiments/breast_cancer/stage1/params.json b/experiments/breast_cancer/stage1/params.json
index 43739c2db86556421d35a2064aa554b532a5b413..d2bca84adc1377209cdc03c17d8a3d965bf233da 100644
--- a/experiments/breast_cancer/stage1/params.json
+++ b/experiments/breast_cancer/stage1/params.json
@@ -3,9 +3,9 @@
"best_score_train": 0.9562271062271059,
"best_score_test": 0.9514619883040936,
"best_parameters": {
- "max_depth": "20",
- "min_samples_leaf": "1",
- "n_estimators": "1000",
+ "max_depth": 20,
+ "min_samples_leaf": 1,
+ "n_estimators": 1000,
"max_features": "log2"
},
"random_seed": [
diff --git a/experiments/california_housing/stage1/params.json b/experiments/california_housing/stage1/params.json
new file mode 100644
index 0000000000000000000000000000000000000000..617c93c1725070111e2ac5a0292ef450c389df3c
--- /dev/null
+++ b/experiments/california_housing/stage1/params.json
@@ -0,0 +1,16 @@
+{
+ "scorer": "neg_mean_squared_error",
+ "best_score_train": -0.2535049905518054,
+ "best_score_test": -0.24128661227361273,
+ "best_parameters": {
+ "max_features": "log2",
+ "min_samples_leaf": 1,
+ "n_estimators": 1000,
+ "max_depth": 18
+ },
+ "random_seed": [
+ 1012,
+ 529,
+ 42
+ ]
+}
\ No newline at end of file
diff --git a/experiments/diabetes/stage1/params.json b/experiments/diabetes/stage1/params.json
index 472e7382583edd3ce381470a7ba9902aff443f5d..2ade87a85a4ae7ca8a30eeea974173163e576848 100644
--- a/experiments/diabetes/stage1/params.json
+++ b/experiments/diabetes/stage1/params.json
@@ -4,9 +4,9 @@
"best_score_test": -3305.635542701523,
"best_parameters": {
"max_features": "auto",
- "min_samples_leaf": "1",
- "max_depth": "15",
- "n_estimators": "108"
+ "min_samples_leaf": 1,
+ "max_depth": 15,
+ "n_estimators": 108
},
"random_seed": [
661,
diff --git a/experiments/digits/stage1/params.json b/experiments/digits/stage1/params.json
index 2cf1c4b1f5cc91bda737fe2f68c81d37d2682f2a..845a1aec4ffdb77db0d9eda6ca712dbb80c5bfb9 100644
--- a/experiments/digits/stage1/params.json
+++ b/experiments/digits/stage1/params.json
@@ -4,9 +4,9 @@
"best_score_test": 0.9738888888888889,
"best_parameters": {
"max_features": "sqrt",
- "min_samples_leaf": "1",
- "n_estimators": "1000",
- "max_depth": "20"
+ "min_samples_leaf": 1,
+ "n_estimators": 1000,
+ "max_depth": 20
},
"random_seed": [
1,
diff --git a/experiments/iris/stage1/params.json b/experiments/iris/stage1/params.json
index a91c658dfdf58b7338807194dbbe1f01b70aa431..e3eacac482d4747992a15987eaaabc54ac23c13b 100644
--- a/experiments/iris/stage1/params.json
+++ b/experiments/iris/stage1/params.json
@@ -4,9 +4,9 @@
"best_score_test": 0.9155555555555556,
"best_parameters": {
"max_features": "sqrt",
- "min_samples_leaf": "1",
- "max_depth": "1",
- "n_estimators": "1000"
+ "min_samples_leaf": 1,
+ "max_depth": 1,
+ "n_estimators": 1000
},
"random_seed": [
771,
diff --git a/experiments/iris/stage1/with_best_params_16.json b/experiments/iris/stage1/with_best_params_16.json
new file mode 100644
index 0000000000000000000000000000000000000000..102999a98070b312cdb6c17d7767423fe4ab75b7
--- /dev/null
+++ b/experiments/iris/stage1/with_best_params_16.json
@@ -0,0 +1,36 @@
+{
+ "experiment_configuration": null,
+ "experiment_configuration_path": "experiments",
+ "dataset_name": "iris",
+ "normalize_D": false,
+ "dataset_normalizer": "standard",
+ "forest_size": null,
+ "extracted_forest_size_samples": 4,
+ "models_dir": ".\\models",
+ "dev_size": 0.2,
+ "test_size": 0.2,
+ "random_seed_number": 1,
+ "seeds": [
+ 1,
+ 2,
+ 3,
+ 4,
+ 5
+ ],
+ "subsets_used": "train,dev",
+ "normalize_weights": false,
+ "verbose": false,
+ "skip_best_hyperparams": false,
+ "save_experiment_configuration": [
+ "1",
+ "with_best_params"
+ ],
+ "job_number": -1,
+ "extracted_forest_size": [
+ 200,
+ 400,
+ 600,
+ 800
+ ],
+ "experiment_id": 16
+}
\ No newline at end of file
diff --git a/experiments/iris/stage1/wo_best_params_17.json b/experiments/iris/stage1/wo_best_params_17.json
new file mode 100644
index 0000000000000000000000000000000000000000..0294d64c0f98e3399e1b76b544204b821196c2fb
--- /dev/null
+++ b/experiments/iris/stage1/wo_best_params_17.json
@@ -0,0 +1,36 @@
+{
+ "experiment_configuration": null,
+ "experiment_configuration_path": "experiments",
+ "dataset_name": "iris",
+ "normalize_D": false,
+ "dataset_normalizer": "standard",
+ "forest_size": null,
+ "extracted_forest_size_samples": 4,
+ "models_dir": ".\\models",
+ "dev_size": 0.2,
+ "test_size": 0.2,
+ "random_seed_number": 1,
+ "seeds": [
+ 1,
+ 2,
+ 3,
+ 4,
+ 5
+ ],
+ "subsets_used": "train,dev",
+ "normalize_weights": false,
+ "verbose": false,
+ "skip_best_hyperparams": true,
+ "save_experiment_configuration": [
+ "1",
+ "wo_best_params"
+ ],
+ "job_number": -1,
+ "extracted_forest_size": [
+ 20,
+ 40,
+ 60,
+ 80
+ ],
+ "experiment_id": 17
+}
\ No newline at end of file
diff --git a/experiments/linnerud/stage1/params.json b/experiments/linnerud/stage1/params.json
index a1573d389eb93336ada91ee31ac66a7de166cd33..7e45dc06e3b617f4fb98c6645b9d3a193ba9d44b 100644
--- a/experiments/linnerud/stage1/params.json
+++ b/experiments/linnerud/stage1/params.json
@@ -3,10 +3,10 @@
"best_score_train": -223.81438159498393,
"best_score_test": -262.4415311793658,
"best_parameters": {
- "max_depth": "1",
- "min_samples_leaf": "1",
+ "max_depth": 1,
+ "min_samples_leaf": 1,
"max_features": "sqrt",
- "n_estimators": "1000"
+ "n_estimators": 1000
},
"random_seed": [
1109,
diff --git a/experiments/olivetti_faces/stage1/params.json b/experiments/olivetti_faces/stage1/params.json
new file mode 100644
index 0000000000000000000000000000000000000000..c9d83bfd06cf6a9f18bb319538626b322f5b5b7b
--- /dev/null
+++ b/experiments/olivetti_faces/stage1/params.json
@@ -0,0 +1,28 @@
+{
+ "scorer": "accuracy",
+ "best_score_train": 0.8890625,
+ "best_score_test": 0.89,
+ "best_parameters": {
+ "max_features": "log2",
+ "min_samples_leaf": 1,
+ "n_estimators": 1000,
+ "max_depth": 18
+ },
+ "random_seed": [
+ 899,
+ 249,
+ 1367,
+ 942,
+ 846,
+ 1576,
+ 285,
+ 839,
+ 1974,
+ 1216,
+ 540,
+ 1292,
+ 1642,
+ 712,
+ 1511
+ ]
+}
\ No newline at end of file
diff --git a/experiments/wine/stage1/params.json b/experiments/wine/stage1/params.json
index 99795a6caf3b2c0639df4d0bc3306f0906193309..2ede59c2abaa9ee572d84cd256eedb226ac80773 100644
--- a/experiments/wine/stage1/params.json
+++ b/experiments/wine/stage1/params.json
@@ -3,9 +3,9 @@
"best_score_train": 0.9846607669616517,
"best_score_test": 0.9796296296296295,
"best_parameters": {
- "max_depth": "20",
- "min_samples_leaf": "1",
- "n_estimators": "1000",
+ "max_depth": 20,
+ "min_samples_leaf": 1,
+ "n_estimators": 1000,
"max_features": "log2"
},
"random_seed": [