[Refactoring] Put all utility routines within one package. (#196)

* utils refactoring;


Co-authored-by: Yuriy Romanyshyn <yuriy.romanyshyn.lv.ua@gmail.com>

deepchecks/__init__.py

@@ -3,7 +3,8 @@
 import matplotlib
 import matplotlib.pyplot as plt
 
-from .utils import is_notebook
+from .utils.ipython import is_notebook
+
 # Matplotlib has multiple backends. If we are in a context that does not support GUI (For example, during unit tests)
 # we can't use a GUI backend. Thus we must use a non-GUI backend.
 if not is_notebook():

deepchecks/base/__init__.py

@@ -5,3 +5,4 @@
 from .dataset import *
 from .check import *
 from .suite import *
+

deepchecks/base/check.py

@@ -15,8 +15,8 @@
 from pandas.io.formats.style import Styler
 
 from deepchecks.base.display_pandas import display_dataframe
-from deepchecks.string_utils import split_camel_case
-from deepchecks.utils import DeepchecksValueError
+from deepchecks.utils.strings import split_camel_case
+from deepchecks.errors import DeepchecksValueError
 
 
 class Condition:

deepchecks/base/dataset.py

@@ -4,12 +4,14 @@
 import pandas as pd
 from pandas.core.dtypes.common import is_float_dtype
 
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.utils import DeepchecksValueError
-from deepchecks.string_utils import is_string_column
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.errors import DeepchecksValueError
+from deepchecks.utils.strings import is_string_column
+
 
 __all__ = ['Dataset', 'ensure_dataframe_type']
 
+
 logger = logging.getLogger('deepchecks.dataset')
 
 

deepchecks/base/display_suite.py

@@ -8,7 +8,7 @@
 
 from deepchecks.base.check import CheckResult, CheckFailure
 from deepchecks.base.display_pandas import dataframe_to_html, display_dataframe
-from deepchecks.utils import is_widgets_enabled
+from deepchecks.utils.ipython import is_widgets_enabled
 import pandas as pd
 
 __all__ = ['display_suite_result', 'ProgressBar']

deepchecks/base/suite.py

@@ -4,7 +4,7 @@
 from typing import Union, List, Optional
 
 from deepchecks.base.display_suite import display_suite_result, ProgressBar
-from deepchecks.utils import DeepchecksValueError
+from deepchecks.errors import DeepchecksValueError
 from deepchecks.base import Dataset
 from deepchecks.base.check import (
     BaseCheck, CheckResult, TrainTestBaseCheck, CompareDatasetsBaseCheck,

deepchecks/checks/distribution/train_test_drift.py

@@ -9,8 +9,8 @@
 
 from deepchecks import Dataset, CheckResult, TrainTestBaseCheck, ConditionResult
 from deepchecks.checks.distribution.plot import plot_density
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null
-from deepchecks.utils import DeepchecksValueError
+from deepchecks.utils.features import calculate_feature_importance_or_null
+from deepchecks.errors import DeepchecksValueError
 import matplotlib.pyplot as plt
 
 __all__ = ['TrainTestDrift']

deepchecks/checks/distribution/trust_score_comparison.py

@@ -6,9 +6,11 @@
 from deepchecks.checks.distribution.trust_score import TrustScore
 from deepchecks.checks.distribution.preprocessing import preprocess_dataset_to_scaled_numerics
 from deepchecks.checks.distribution.plot import plot_density
-from deepchecks.metric_utils import task_type_check, ModelType
-from deepchecks.string_utils import format_percent
-from deepchecks.utils import DeepchecksValueError, model_type_validation
+from deepchecks.utils.metrics import task_type_check, ModelType
+from deepchecks.utils.strings import format_percent
+from deepchecks.utils.validation import model_type_validation
+from deepchecks.errors import DeepchecksValueError
+
 
 __all__ = ['TrustScoreComparison']
 

deepchecks/checks/integrity/data_duplicates.py

@@ -5,9 +5,9 @@
 
 from deepchecks import Dataset, ensure_dataframe_type
 from deepchecks.base.check import CheckResult, SingleDatasetBaseCheck, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.utils import DeepchecksValueError
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.strings import format_percent
+from deepchecks.errors import DeepchecksValueError
 
 
 __all__ = ['DataDuplicates']

deepchecks/checks/integrity/dominant_frequency_change.py

@@ -7,12 +7,12 @@
 
 from deepchecks import Dataset
 from deepchecks.base.check import CheckResult, CompareDatasetsBaseCheck, ConditionResult
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import format_percent
+from deepchecks.errors import DeepchecksValueError
 
-__all__ = ['DominantFrequencyChange']
 
-from deepchecks.string_utils import format_percent
-from deepchecks.utils import DeepchecksValueError
+__all__ = ['DominantFrequencyChange']
 
 
 class DominantFrequencyChange(CompareDatasetsBaseCheck):

deepchecks/checks/integrity/is_single_value.py

@@ -2,8 +2,9 @@
 from typing import Union, Iterable
 import pandas as pd
 from deepchecks import SingleDatasetBaseCheck, CheckResult, ensure_dataframe_type, Dataset, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.string_utils import format_columns_for_condition
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.strings import format_columns_for_condition
+
 
 __all__ = ['IsSingleValue']
 

deepchecks/checks/integrity/label_ambiguity.py

@@ -5,7 +5,7 @@
 
 from deepchecks import Dataset, ConditionResult
 from deepchecks.base.check import CheckResult, SingleDatasetBaseCheck
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.strings import format_percent
 
 
 __all__ = ['LabelAmbiguity']

deepchecks/checks/integrity/mixed_nulls.py

@@ -7,14 +7,15 @@
 
 from deepchecks import Dataset, CheckResult, ensure_dataframe_type
 from deepchecks.base.check import SingleDatasetBaseCheck, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import string_baseform, format_percent, format_columns_for_condition
+from deepchecks.errors import DeepchecksValueError
 
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
-from deepchecks.string_utils import string_baseform, format_percent, format_columns_for_condition
-from deepchecks.utils import DeepchecksValueError
 
 __all__ = ['MixedNulls']
 
+
 DEFAULT_NULL_VALUES = {'none', 'null', 'nan', 'na', '', '\x00', '\x00\x00'}
 
 

deepchecks/checks/integrity/mixed_types.py

@@ -6,14 +6,13 @@
 
 from deepchecks import Dataset, ensure_dataframe_type
 from deepchecks.base.check import CheckResult, SingleDatasetBaseCheck, ConditionResult
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import is_string_column, format_percent, format_columns_for_condition
 
 
 __all__ = ['MixedTypes']
 
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.string_utils import is_string_column, format_percent, format_columns_for_condition
-
 
 class MixedTypes(SingleDatasetBaseCheck):
     """Search for various types of data in (a) column[s], including hidden mixes in strings.

deepchecks/checks/integrity/new_category.py

@@ -1,10 +1,11 @@
 """The data_sample_leakage_report check module."""
 from typing import Union, Iterable, Dict
+import pandas as pd
+
 from deepchecks import Dataset
 from deepchecks.base.check import CheckResult, TrainTestBaseCheck, ConditionResult
-from deepchecks.string_utils import format_percent, format_columns_for_condition
+from deepchecks.utils.strings import format_percent, format_columns_for_condition
 
-import pandas as pd
 
 __all__ = ['CategoryMismatchTrainTest']
 

deepchecks/checks/integrity/new_label.py

@@ -3,12 +3,13 @@
 
 from deepchecks import Dataset
 from deepchecks.base.check import CheckResult, TrainTestBaseCheck, ConditionResult
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.strings import format_percent
 
 import pandas as pd
 
 pd.options.mode.chained_assignment = None
 
+
 __all__ = ['NewLabelTrainTest']
 
 

deepchecks/checks/integrity/rare_format_detection.py

@@ -6,11 +6,11 @@
 import pandas as pd
 
 from deepchecks import CheckResult, Dataset, SingleDatasetBaseCheck, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
 from deepchecks.base.dataset import ensure_dataframe_type
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_dict
-from deepchecks.string_utils import split_and_keep, split_by_order, format_percent
-from deepchecks.utils import DeepchecksValueError
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_dict
+from deepchecks.utils.strings import split_and_keep, split_by_order, format_percent
+from deepchecks.errors import DeepchecksValueError
 
 
 __all__ = ['RareFormatDetection', 'Pattern']

deepchecks/checks/integrity/special_chars.py

@@ -6,9 +6,10 @@
 
 from deepchecks import Dataset, ensure_dataframe_type
 from deepchecks.base.check import CheckResult, SingleDatasetBaseCheck, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
-from deepchecks.string_utils import string_baseform, format_percent, format_columns_for_condition
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import string_baseform, format_percent, format_columns_for_condition
+
 
 __all__ = ['SpecialCharacters']
 

deepchecks/checks/integrity/string_length_out_of_bounds.py

@@ -9,9 +9,10 @@
 from scipy import stats
 
 from deepchecks import CheckResult, SingleDatasetBaseCheck, Dataset, ensure_dataframe_type, ConditionResult
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
-from deepchecks.string_utils import is_string_column, format_number, format_columns_for_condition, format_percent
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import is_string_column, format_number, format_columns_for_condition, format_percent
+from deepchecks.utils.dataframes import filter_columns_with_validation
+
 
 __all__ = ['StringLengthOutOfBounds']
 

deepchecks/checks/integrity/string_mismatch.py

@@ -4,12 +4,22 @@
 
 import pandas as pd
 
-from deepchecks import CheckResult, SingleDatasetBaseCheck, Dataset, ensure_dataframe_type, ConditionResult, \
+from deepchecks import (
+    CheckResult,
+    SingleDatasetBaseCheck,
+    Dataset,
+    ensure_dataframe_type,
+    ConditionResult,
     ConditionCategory
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.string_utils import get_base_form_to_variants_dict, is_string_column, format_percent, \
+)
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.strings import (
+    get_base_form_to_variants_dict,
+    is_string_column,
+    format_percent,
     format_columns_for_condition
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
+)
 
 
 __all__ = ['StringMismatch']

deepchecks/checks/integrity/string_mismatch_comparison.py

@@ -5,10 +5,11 @@
 import pandas as pd
 
 from deepchecks import CheckResult, Dataset, ensure_dataframe_type, CompareDatasetsBaseCheck, ConditionResult
-from deepchecks.base.dataframe_utils import filter_columns_with_validation
-from deepchecks.string_utils import get_base_form_to_variants_dict, is_string_column, format_percent, \
+from deepchecks.utils.dataframes import filter_columns_with_validation
+from deepchecks.utils.strings import get_base_form_to_variants_dict, is_string_column, format_percent, \
     format_columns_for_condition
-from deepchecks.feature_importance_utils import calculate_feature_importance_or_null, column_importance_sorter_df
+from deepchecks.utils.features import calculate_feature_importance_or_null, column_importance_sorter_df
+
 
 __all__ = ['StringMismatchComparison']
 

deepchecks/checks/methodology/boosting_overfit.py

@@ -7,9 +7,10 @@
 import numpy as np
 
 from deepchecks import Dataset, CheckResult, TrainTestBaseCheck, ConditionResult
-from deepchecks.metric_utils import task_type_check, DEFAULT_METRICS_DICT, validate_scorer, DEFAULT_SINGLE_METRIC
-from deepchecks.string_utils import format_percent
-from deepchecks.utils import DeepchecksValueError
+from deepchecks.utils.metrics import task_type_check, DEFAULT_METRICS_DICT, validate_scorer, DEFAULT_SINGLE_METRIC
+from deepchecks.utils.strings import format_percent
+from deepchecks.errors import DeepchecksValueError
+
 
 __all__ = ['BoostingOverfit']
 

deepchecks/checks/methodology/date_train_test_leakage_duplicates.py

@@ -2,7 +2,8 @@
 import pandas as pd
 
 from deepchecks import CheckResult, Dataset, TrainTestBaseCheck, ConditionResult
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.strings import format_percent
+
 
 __all__ = ['DateTrainTestLeakageDuplicates']
 

deepchecks/checks/methodology/date_train_test_leakage_overlap.py

@@ -1,6 +1,7 @@
 """The date_leakage check module."""
 from deepchecks import CheckResult, Dataset, TrainTestBaseCheck, ConditionResult
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.strings import format_percent
+
 
 __all__ = ['DateTrainTestLeakageOverlap']
 

deepchecks/checks/methodology/identifier_leakage.py

@@ -3,12 +3,12 @@
 
 import pandas as pd
 
+import deepchecks.ppscore as pps
 from deepchecks import Dataset
 from deepchecks.base.check import CheckResult, SingleDatasetBaseCheck, ConditionResult
-from deepchecks.plot_utils import create_colorbar_barchart_for_check
-from deepchecks.string_utils import format_percent
-from deepchecks.utils import DeepchecksValueError
-import deepchecks.ppscore as pps
+from deepchecks.utils.plot import create_colorbar_barchart_for_check
+from deepchecks.utils.strings import format_percent
+from deepchecks.errors import DeepchecksValueError
 
 
 __all__ = ['IdentifierLeakage']

deepchecks/checks/methodology/index_leakage.py

@@ -3,7 +3,8 @@
 
 from deepchecks import CheckResult, Dataset, TrainTestBaseCheck
 from deepchecks.base.check import ConditionResult
-from deepchecks.string_utils import format_percent
+from deepchecks.utils.strings import format_percent
+
 
 __all__ = ['IndexTrainTestLeakage']
 

deepchecks/checks/methodology/performance_overfit.py

@@ -5,9 +5,9 @@
 import pandas as pd
 import numpy as np
 
-from deepchecks.string_utils import format_percent
-from deepchecks.utils import model_type_validation
-from deepchecks.metric_utils import get_metrics_list
+from deepchecks.utils.strings import format_percent
+from deepchecks.utils.validation import model_type_validation
+from deepchecks.utils.metrics import get_metrics_list
 from deepchecks import (
     Dataset,
     CheckResult,

deepchecks/checks/methodology/single_feature_contribution.py

@@ -2,7 +2,7 @@
 import typing as t
 
 import deepchecks.ppscore as pps
-from deepchecks.plot_utils import create_colorbar_barchart_for_check
+from deepchecks.utils.plot import create_colorbar_barchart_for_check
 from deepchecks import CheckResult, Dataset, SingleDatasetBaseCheck, ConditionResult
 
 

deepchecks/checks/methodology/single_feature_contribution_train_validation.py

@@ -3,7 +3,7 @@
 
 import deepchecks.ppscore as pps
 from deepchecks import CheckResult, Dataset, TrainTestBaseCheck, ConditionResult
-from deepchecks.plot_utils import create_colorbar_barchart_for_check
+from deepchecks.utils.plot import create_colorbar_barchart_for_check