The Utilities Space

Helper cells shared by Assumptions and PolicyAttrs.

This Space is used as a base Space (_bases) for Assumptions and PolicyAttrs, which both derive their per-policy values by mapping pandas tables onto the rows of the policy data.

The cells in this Space rely on a return_array reference, which is defined by the inheriting space (True to convert results to NumPy arrays, False to keep them as pandas objects).

Cells Summary

Helpers that convert pandas objects to per-policy NumPy arrays and reindex assumption tables onto the policy data.

pandas_to_array(df_or_series)	Return df_or_series as a NumPy array or as is.
map_to_policies(series)	Reindex an assumption series to one entry per policy.

Cells Descriptions

pandas_to_array(df_or_series)

Return df_or_series as a NumPy array or as is.

If the inheriting space sets return_array = True (the default for Assumptions and PolicyAttrs), the input is converted via to_numpy(). Otherwise the original pandas object is returned unchanged.

This is an uncached cell, so it is recomputed on every call.

map_to_policies(series)

Reindex an assumption series to one entry per policy.

Looks up the values of series for each row of policy_data() using the columns identified by series.index.names (e.g. Product, PolType, Gen), and returns a Series whose index matches policy_data so it aligns with the per-policy NumPy arrays used elsewhere in the model.

This is an uncached cell, so it is recomputed on every call.