EdaValidator

Calculates EDA statistics for numeric columns, returns them in result.stats and, by default, prints them as a Rich table.

Basic Usage

from epftoolbox2.data.validators import EdaValidator

validator = EdaValidator()
result = validator.validate(df)

quiet = EdaValidator(columns=["price", "load_actual"], verbose=False)
stats_df = quiet.validate(df).stats

Parameters

Parameter	Type	Default	Description
columns	List[str] or None	None	Columns to analyse. None uses every numeric column.
verbose	bool	True	Print the statistics table to the console. Set False for unattended runs.

Console Output

┏━━━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━┓
┃ Column      ┃ Min      ┃ Max      ┃ Mean     ┃ Std      ┃
┡━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━┩
│ load_actual │ 8500.12  │ 25000.45 │ 15234.67 │ 3456.78  │
│ price       │ -5.23    │ 450.00   │ 85.45    │ 45.23    │
└─────────────┴──────────┴──────────┴──────────┴──────────┘

Statistics Included

Statistic	Description
Min	Minimum value
Max	Maximum value
Mean	Average value
Std	Standard deviation
Q25	25th percentile
Q50	Median (50th percentile)
Q75	75th percentile
Nulls	Count of null values

Accessing Statistics Programmatically

result = validator.validate(df)

# Get stats as DataFrame
stats_df = result.stats
print(stats_df)

When to Use

• Initial data exploration
• Checking for outliers
• Verifying data ranges are reasonable (for automated range checks use PlausibilityValidator)