Metrics data within the system is managed in a three-level hierarchical format: Category -> Usage > Source, represented as the three values separated by a . character. For example, we store Category=Energy, Usage=Heating, Source=Oil as Energy.Heating.Oil. For data where we want to demonstrate the totality of a combination, we expect a blank value in the source data and use an ALL stand-in internally (i.e. Category=Energy, Usage=Heating, Source= represents all heating energy and would be represented by Energy.Heating.ALL).
However, we've noticed in the Municipal Data source that some data exists outside of the expected hierarchy. For example, there are values where the Source is populated but not the Usage (e.g. Energy.ALL.Electricity).
Additionally, there are situations where we have categories but no ALL (e.g. Energy.Heating.Oil but no Energy.Heating.ALL).
Neither of these situations are bugs in terms of how we use the data (i.e. our system can handle these situations), however the do represent oddities in what one may expect from the data.
Metrics data within the system is managed in a three-level hierarchical format:
Category
->Usage
>Source
, represented as the three values separated by a.
character. For example, we storeCategory=Energy
,Usage=Heating
,Source=Oil
asEnergy.Heating.Oil
. For data where we want to demonstrate the totality of a combination, we expect a blank value in the source data and use anALL
stand-in internally (i.e.Category=Energy
,Usage=Heating
,Source=
represents all heating energy and would be represented byEnergy.Heating.ALL
).However, we've noticed in the Municipal Data source that some data exists outside of the expected hierarchy. For example, there are values where the Source is populated but not the Usage (e.g.
Energy.ALL.Electricity
).Additionally, there are situations where we have categories but no ALL (e.g.
Energy.Heating.Oil
but noEnergy.Heating.ALL
).Neither of these situations are bugs in terms of how we use the data (i.e. our system can handle these situations), however the do represent oddities in what one may expect from the data.