This is a fix to predictsignauxfaibles.explainfunction explain, which attempts to replace the contribution of micro features from group retards_paiement with the aggregated contribution, so as to have a single variable retards_paiement as an explanation for the contribution of variables coming from Altares.
The function used to attempt this aggregation whether or not the features from retards_paiement existed.
Here we just check that they exist before aggregation.
In the long run, we should create a dictionary of micro variables to group, so as to make generic both function group_retards_paiement and the condition in this fix.
This is a fix to
predictsignauxfaibles.explain
functionexplain
, which attempts to replace the contribution of micro features from groupretards_paiement
with the aggregated contribution, so as to have a single variableretards_paiement
as an explanation for the contribution of variables coming from Altares.The function used to attempt this aggregation whether or not the features from
retards_paiement
existed. Here we just check that they exist before aggregation.In the long run, we should create a dictionary of micro variables to group, so as to make generic both function
group_retards_paiement
and the condition in this fix.