Regression and multi-objective APIs corrections

Correction of the API to resolve issues with colab examples

docs/source/conf.py

@@ -22,7 +22,7 @@ copyright = '2020-2021, NSS Lab'
 author = 'NSS Lab'
 
 # The full version, including alpha/beta/rc tags
-release = '0.2.1'
+release = '0.2.2'
 
 # -- General configuration ---------------------------------------------------
 

examples/fedot_api_example.py

+import pandas as pd
+
 from fedot.api.main import Fedot
+from fedot.core.data.data import InputData, train_test_data_setup
 from fedot.core.utils import project_root
 
 
@@ -11,7 +14,7 @@ def run_classification_example():
     baseline_model = Fedot(problem=problem)
     baseline_model.fit(features=train_data_path, target='target', predefined_model='xgboost')
 
-    baseline_model.predict_proba(features=test_data_path)
+    baseline_model.predict(features=test_data_path)
     print(baseline_model.get_metrics())
 
     auto_model = Fedot(problem=problem, seed=42)
@@ -22,6 +25,28 @@ def run_classification_example():
     return prediction
 
 
+def run_regression_example():
+    data_path = f'{project_root()}/cases/data/cholesterol/cholesterol.csv'
+
+    data = InputData.from_csv(data_path)
+    train, test = train_test_data_setup(data)
+
+    problem = 'regression'
+
+    baseline_model = Fedot(problem=problem)
+    baseline_model.fit(features=train, predefined_model='xgbreg')
+
+    baseline_model.predict(features=test)
+    print(baseline_model.get_metrics())
+
+    auto_model = Fedot(problem=problem, seed=42)
+    auto_model.fit(features=train, target='target')
+    prediction = auto_model.predict(features=test)
+    print(auto_model.get_metrics())
+
+    return prediction
+
+
 def run_ts_forecasting_example(with_plot=True):
     train_data_path = f'{project_root()}/notebooks/jupyter_media/intro/salaries.csv'
 
@@ -29,9 +54,13 @@ def run_ts_forecasting_example(with_plot=True):
     model = Fedot(problem='ts_forecasting')
 
     # run AutoML model design in the same way
-    chain = model.fit(features=train_data_path, target='target')
+    chain = model.fit(features=train_data_path, target='target', )
     chain.show()
 
+    model.predict(features=train_data_path)
+
+    print(model.get_metrics())
+
     # use model to obtain forecast
     forecast = model.forecast(pre_history=train_data_path, forecast_length=30)
 
@@ -42,7 +71,27 @@ def run_ts_forecasting_example(with_plot=True):
     return forecast
 
 
+def run_classification_multiobj_example(with_plot=True):
+    train_data = pd.read_csv(f'./data/Hill_Valley_with_noise_Training.data')
+    test_data = pd.read_csv(f'./data/Hill_Valley_with_noise_Testing.data')
+    target = test_data['class']
+    del test_data['class']
+    problem = 'classification'
+
+    auto_model = Fedot(problem=problem, learning_time=5, preset='light',
+                       composer_params={'metric': ['f1', 'node_num']}, seed=42)
+    auto_model.fit(features=train_data, target='class')
+    prediction = auto_model.predict_proba(features=test_data)
+    print(auto_model.get_metrics(target))
+
+    if with_plot:
+        auto_model.best_models.show()
+
+    return prediction
+
+
 if __name__ == '__main__':
     run_classification_example()
-
-    run_ts_forecasting_example()
+run_classification_multiobj_example()
+run_regression_example()
+run_ts_forecasting_example()

fedot/api/api_utils.py

@@ -10,8 +10,9 @@ from fedot.core.repository.dataset_types import DataTypesEnum
 from fedot.core.repository.model_types_repository import (
     ModelTypesRepository
 )
-from fedot.core.repository.quality_metrics_repository import ClassificationMetricsEnum, \
-    ClusteringMetricsEnum, RegressionMetricsEnum, MetricsRepository
+from fedot.core.repository.quality_metrics_repository import (ClassificationMetricsEnum, ClusteringMetricsEnum,
+                                                              ComplexityMetricsEnum, MetricsRepository,
+                                                              RegressionMetricsEnum)
 from fedot.core.repository.tasks import Task, TaskTypesEnum
 from fedot.utilities.define_metric_by_task import MetricByTask
 
@@ -23,13 +24,15 @@ metrics_mapping = {
     'mae': RegressionMetricsEnum.MAE,
     'mse': RegressionMetricsEnum.MSE,
     'msle': RegressionMetricsEnum.MSLE,
+    'mape': RegressionMetricsEnum.MAPE,
     'r2': RegressionMetricsEnum.R2,
     'rmse': RegressionMetricsEnum.RMSE,
-    'silhouette': ClusteringMetricsEnum.silhouette
+    'silhouette': ClusteringMetricsEnum.silhouette,
+    'node_num': ComplexityMetricsEnum.node_num
 }
 
 
-def autodetect_data_type(task: Task) -> DataTypesEnum:
+def _autodetect_data_type(task: Task) -> DataTypesEnum:
     if task.task_type == TaskTypesEnum.ts_forecasting:
         return DataTypesEnum.ts
     else:
@@ -48,7 +51,7 @@ def save_predict(predicted_data: OutputData):
 def array_to_input_data(features_array: np.array,
                         target_array: np.array,
                         task: Task = Task(TaskTypesEnum.classification)):
-    data_type = autodetect_data_type(task)
+    data_type = _autodetect_data_type(task)
     idx = np.arange(len(features_array))
 
     return InputData(idx=idx, features=features_array, target=target_array, task=task, data_type=data_type)
@@ -78,7 +81,7 @@ def compose_fedot_model(train_data: InputData,
                         max_arity: int,
                         pop_size: int,
                         num_of_generations: int,
-                        learning_time: int = 5,
+                        learning_time: float = 5,
                         available_model_types: list = None,
                         with_tuning=False,
                         metric=None
@@ -87,10 +90,16 @@ def compose_fedot_model(train_data: InputData,
     if metric is None:
         metric_function = MetricByTask(task.task_type).metric_cls.get_value
     else:
-        metric_id = metrics_mapping.get(metric, None)
-        if metric_id is None:
-            raise ValueError(f'Incorrect metric {metric}')
-        metric_function = MetricsRepository().metric_by_id(metric_id)
+        if isinstance(metric, str):
+            metric = [metric]
+
+        metric_function = []
+        for specific_metric in metric:
+            metric_id = metrics_mapping.get(specific_metric, None)
+            if metric_id is None:
+                raise ValueError(f'Incorrect metric {specific_metric}')
+            specific_metric_function = MetricsRepository().metric_by_id(metric_id)
+            metric_function.append(specific_metric_function)
 
     learning_time = datetime.timedelta(minutes=learning_time)
 
@@ -114,12 +123,20 @@ def compose_fedot_model(train_data: InputData,
 
     logger.message('Model composition started')
     chain_gp_composed = gp_composer.compose_chain(data=train_data)
-    chain_gp_composed.log = logger
+
+    chain_for_tune = chain_gp_composed
+    chain_for_return = chain_gp_composed
+
+    if isinstance(chain_gp_composed, list):
+        for chain in chain_gp_composed:
+            chain.log = logger
+        chain_for_tune = chain_gp_composed[0]
+        chain_for_return = gp_composer.optimiser.archive
 
     if with_tuning:
         logger.message('Hyperparameters tuning started')
-        chain_gp_composed.fine_tune_primary_nodes(input_data=train_data)
+        chain_for_tune.fine_tune_primary_nodes(input_data=train_data)
 
     logger.message('Model composition finished')
 
-    return chain_gp_composed
+    return chain_for_return

[Elizaveta Lutsenko]

Created by: pep8speaks

Hello @nicl-nno! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers!

Comment last updated at 2021-04-02 15:16:16 UTC

[Elizaveta Lutsenko]

Created by: codecov[bot]

Codecov Report

Merging #257 (fc124458) into master (00d64c41) will decrease coverage by 0.06%. The diff coverage is 83.33%.

@@            Coverage Diff             @@
##           master     #257      +/-   ##
==========================================
- Coverage   77.87%   77.80%   -0.07%     
==========================================
  Files          71       72       +1     
  Lines        5391     5438      +47     
==========================================
+ Hits         4198     4231      +33     
- Misses       1193     1207      +14     

Impacted Files	Coverage Δ
fedot/core/composer/optimisers/utils/pareto.py	66.66% <66.66%> (ø)
...core/composer/optimisers/utils/population_utils.py	78.57% <66.66%> (ø)
fedot/core/composer/visualisation.py	30.58% <66.66%> (-0.24%)
.../composer/optimisers/gp_comp/operators/mutation.py	66.26% <77.77%> (-2.41%)
fedot/api/main.py	82.65% <80.00%> (-0.39%)
fedot/api/api_utils.py	86.76% <85.71%> (+0.79%)
...composer/optimisers/gp_comp/operators/crossover.py	77.08% <87.50%> (ø)
fedot/core/chains/chain.py	96.12% <100.00%> (ø)
fedot/core/chains/chain_template.py	93.70% <100.00%> (+0.04%)
fedot/core/composer/composing_history.py	72.91% <100.00%> (ø)
... and 12 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 00d64c4...fc12445. Read the comment docs.