Integrate RAStock data #20
Conversation
| @@ -953,6 +955,8 @@ def transform_building_features() -> DataFrame: | |||
| # COMMAND ---------- | |||
|
|
|||
| # DBTITLE 1,Apply upgrade functions | |||
| # TODO: put this in some kind of shared config that can be used across srcipts/repos | |||
| SUPPORTED_UPGRADES = [0.0, 1.0, 3.0, 4.0, 6.0, 9.0, 11.05, 13.01] | |||
There was a problem hiding this comment.
Do we perhaps want to have individual constants for these numbers. I guess or people who work with these every single day, they know what 3 and 7 mean, but for the rest of us, constants would be amazing.
There was a problem hiding this comment.
@nweires where are the enums that you and @christine-george are gonna use gonna live? Can import thsoe use those here?
There was a problem hiding this comment.
I'm still working out exactly what this will look like, but yes, there should be something better to use soon! I'm hoping to move away from using these upgrade ID numbers as the key to identify upgrades.
Draft changes are here - see upgrades.json and the schema in read_upgrades.py.
There was a problem hiding this comment.
thanks Natalie! I will integrate these enums in a subsequent PR when that is ready to roll!
|
|
||
| This function is getting revamped in a different repo so not bothering to document for now | ||
| """ | ||
| # TODO: update with new code from natalie/mohammad in pep |
There was a problem hiding this comment.
This changed already in https://github.com/rewiringamerica/pep/pull/187, but if you want to wait for it to finish changing, this is fine for now.
There was a problem hiding this comment.
yes, i saw that but I thought I'd wait till the upgrade id table was ready and the cchp schema changes were in and then make all the changes in one go.
| @@ -953,6 +955,8 @@ def transform_building_features() -> DataFrame: | |||
| # COMMAND ---------- | |||
|
|
|||
| # DBTITLE 1,Apply upgrade functions | |||
| # TODO: put this in some kind of shared config that can be used across srcipts/repos | |||
| SUPPORTED_UPGRADES = [0.0, 1.0, 3.0, 4.0, 6.0, 9.0, 11.05, 13.01] | |||
There was a problem hiding this comment.
I'm still working out exactly what this will look like, but yes, there should be something better to use soon! I'm hoping to move away from using these upgrade ID numbers as the key to identify upgrades.
Draft changes are here - see upgrades.json and the schema in read_upgrades.py.
Co-authored-by: mikivee <mikivee>
Summary
Adds selected RAStock upgrades to the surrogate model training set. This does NOT retrain the model. Ticket
Detailed Changes
util.pyandconstants.pyscript in preparation for a move to dmutils. Note that some of this code will change after latest changes are merged into PEP.extract_datato use the functions in (1) to pull in all RAStock annual outputs and combine with ResStock outputs.build_feature_store.pydatagen.pyto make sure we train on only subset of building ids, since we don't want the train on the new upgrades for right now.n_building_unitsfeature inbuild_feature_store.pyupdated from str to int.Next
Move util and constants and tests over to dmutils and once this repo is fully migrated over, delete from here and import from there instead.
Who Reviews