Currently, the NetworkReductionData object is created and populated during matrix construction, and we later call populate_branch_maps_by_type! which adds additional maps that are needed for construction of the models in POM.
We can further refine the order of operations such that the NetworkReductionData is computed once with all the common data and mappings that are required in either POM or PF. Note that many of the mappings that are currently only consumed by POM, will likely be required in PF as well during the effort to standardize the POM and PF outputs schema (this will involve the PF outputs looking more like what already exists in POM).
The fields in NetworkReductionData that depend explicitly on the filters applied in the POM template to only model certain lines should be moved to POM, and not stored in the NetworkReductionData object.
This improvement will address the concerns raised in this comment: #233 (comment)
Closing this issue will require cross-package changes in POM/PNM/PF
Currently, the
NetworkReductionDataobject is created and populated during matrix construction, and we later callpopulate_branch_maps_by_type!which adds additional maps that are needed for construction of the models in POM.We can further refine the order of operations such that the NetworkReductionData is computed once with all the common data and mappings that are required in either POM or PF. Note that many of the mappings that are currently only consumed by POM, will likely be required in PF as well during the effort to standardize the POM and PF outputs schema (this will involve the PF outputs looking more like what already exists in POM).
The fields in NetworkReductionData that depend explicitly on the filters applied in the POM template to only model certain lines should be moved to POM, and not stored in the NetworkReductionData object.
This improvement will address the concerns raised in this comment: #233 (comment)
Closing this issue will require cross-package changes in POM/PNM/PF