diagnostics.validate_simulation

validate_simulation(df, input_cols, outcome_col)

Validates simulation data, coercing to numeric and removing invalid rows.

Parameters

Name	Type	Description	Default
df	pd.DataFrame	The raw dataframe containing input columns and the outcome column.	required
input_cols	List[str]	List of input variable names.	required
outcome_col	str	Name of the outcome variable.	required

Returns

Name	Type	Description
	Tuple[pd.DataFrame, pd.DataFrame]	* `df_clean`: The validated, numeric dataframe ready for analysis. * `df_removed`: A dataframe containing the rows that were dropped.

Raises

Name	Type	Description
	ValidationError	If columns are missing, types are wrong, or too few valid rows remain.

Examples

import pandas as pd
# Create dirty data (includes text and negative values)
df = pd.DataFrame({
    'Length': [1.0, 'BadValue', 5.0],
    'Signal': [0.5, 0.8, -1.2]
})

# Validate
clean, removed = validate_simulation(df, ['Length'], 'Signal')
print(f"Clean rows: {len(clean)}")
print(f"Removed rows: {len(removed)}")