timeseries_shaper.stats.boolean_stats

 1import pandas as pd
 2from typing import Dict, Union
 3from ..base import Base
 4
 5class BooleanStatistics(Base):
 6    """
 7    Provides class methods to calculate statistics on a boolean column in a pandas DataFrame.
 8    """
 9
10    @classmethod
11    def count_true(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> int:
12        """Returns the count of True values in the boolean column."""
13        return dataframe[column_name].sum()
14
15    @classmethod
16    def count_false(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> int:
17        """Returns the count of False values in the boolean column."""
18        return (dataframe[column_name] == False).sum()
19
20    @classmethod
21    def count_null(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> int:
22        """Returns the count of null (NaN) values in the boolean column."""
23        return dataframe[column_name].isna().sum()
24
25    @classmethod
26    def count_not_null(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> int:
27        """Returns the count of non-null (True or False) values in the boolean column."""
28        return dataframe[column_name].notna().sum()
29
30    @classmethod
31    def true_percentage(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> float:
32        """Returns the percentage of True values in the boolean column."""
33        true_count = cls.count_true(dataframe, column_name)
34        total_count = cls.count_not_null(dataframe, column_name)
35        return (true_count / total_count) * 100 if total_count > 0 else 0.0
36
37    @classmethod
38    def false_percentage(cls, dataframe: pd.DataFrame, column_name: str = 'value_bool') -> float:
39        """Returns the percentage of False values in the boolean column."""
40        false_count = cls.count_false(dataframe, column_name)
41        total_count = cls.count_not_null(dataframe, column_name)
42        return (false_count / total_count) * 100 if total_count > 0 else 0.0
43
44    @classmethod
45    def mode(cls, dataframe: pd.DataFrame, column_name: str) -> bool:
46        """Returns the mode (most common value) of the specified boolean column."""
47        return dataframe[column_name].mode()[0]
48
49    @classmethod
50    def is_balanced(cls, dataframe: pd.DataFrame, column_name: str) -> bool:
51        """Indicates if the distribution is balanced (50% True and False) in the specified boolean column."""
52        true_percentage = dataframe[column_name].mean()
53        return true_percentage == 0.5
54
55    @classmethod
56    def summary_as_dict(cls, dataframe: pd.DataFrame, column_name: str) -> Dict[str, Union[int, float, bool]]:
57        """Returns a summary of boolean statistics for the specified column as a dictionary."""
58        return {
59            'true_count': cls.count_true(dataframe, column_name),
60            'false_count': cls.count_false(dataframe, column_name),
61            'true_percentage': cls.true_percentage(dataframe, column_name),
62            'false_percentage': cls.false_percentage(dataframe, column_name),
63            'mode': cls.mode(dataframe, column_name),
64            'is_balanced': cls.is_balanced(dataframe, column_name)
65        }
66
67    @classmethod
68    def summary_as_dataframe(cls, dataframe: pd.DataFrame, column_name: str) -> pd.DataFrame:
69        """Returns a summary of boolean statistics for the specified column as a DataFrame."""
70        summary_data = cls.summary_as_dict(dataframe, column_name)
71        return pd.DataFrame([summary_data])