Correlation
Correlation(dataset, column_list, filter_condition_dict=None, duckdb_connection=None, decimal_places=6)
¶
Calculate Pearson correlation coefficients between pairs of numeric columns using DuckDB.
This function computes correlations between specified column pairs using DuckDB's built-in correlation function. It supports multiple column pairs and optional filtering conditions. Each correlation is calculated using complete pairs (non-null values) only.
Parameters¶
dataset : Any Input dataset that can be either: - A DataFrame (pandas, polars) or other DuckDB-compatible data structure - A string representing an existing table name in the DuckDB connection
List[Dict[str, str]]
List of single key-value pair dictionaries. Each dictionary should contain one key-value pair where: - key: first column name for correlation - value: second column name for correlation Example: [{'sales': 'profit'}, {'customers': 'returns'}]
Optional[Dict[str, Union[str, int, float]]], optional
Dictionary of filter conditions to apply before calculating correlations. Format: {'column_name': value} Example: {'department': 'IT', 'year': 2023} Defaults to None.
Optional[DuckDBPyConnection], optional
Existing DuckDB connection. If None, a new connection will be created and closed after execution. Defaults to None.
int, optional
Number of decimal places to round the results. Defaults to 6.
Returns¶
List[Dict[str, Union[str, float, dict, None]]] A list of dictionaries containing: - columns (str): Comma-separated pair of column names (e.g., 'sales,profit') - correlation_value (float): Pearson correlation coefficient Special cases: - None: When one or both columns contain all NULL values - 1.0: When both columns are identical or perfectly correlated - -1.0: When columns are perfectly negatively correlated - sample_size (int): Number of complete pairs used in calculation - table_name (str): Name of the analyzed table - execution_timestamp_utc (str): Timestamp of execution in UTC - filter_conditions (dict|None): Applied filter conditions if any
Raises¶
ValueError Raised if: - column_list is empty or not a list - column pairs are not properly formatted - columns don't exist in the dataset - columns are not numeric - filter conditions reference non-existent columns - decimal_places is negative
Source code in src/whistlingduck/analyzers/Correlation.py
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 |
|