MMMSummaryFactory.residuals_over_time#

MMMSummaryFactory.residuals_over_time(hdi_probs=None, frequency=None, dims=None, output_format=None)[source]#

Create residuals-over-time summary DataFrame.

Parameters:
hdi_probssequence of float, optional

HDI probability levels (default: uses factory default)

frequencyFrequency, optional

Time aggregation period (default: None, no aggregation)

dimsdict, optional

Dimension filters, e.g. {"geo": ["CA"]}.

output_format{“pandas”, “polars”}, optional

Output DataFrame format (default: uses factory default)

Returns:
pd.DataFrame or pl.DataFrame

Summary DataFrame with columns:

  • date: Time index

  • mean, median: Residual point estimates

  • abs_error_{prob}_lower/upper: HDI bounds for each prob

  • <custom_dims>: One column per custom dimension when present