Closed MattStammers closed 1 year ago
Thanks @MattStammers .
I think #36 that @ccarenzoIC raised is subtly different and is around the feature engineering aspect.
Moreover, it will also break the current workflow around how missing or erroneous codes that are not in a refset are labelled.
Please see feature_maps.py
for how this is achieved through a combination of the {category:[array of codes]}
dictionaries that appear later in the same module and the refset members data that appears in nhsdd_snomed.py
.
nhsdd_snomed.py
is meant to be autogenerated using nhsdd_generator.py
.
Happy to jump on a quick call this afternoon if it will help.
Don't merge it if it will break the rest of the workflow. If @ccarenzoIC is also working on this then prioritize what she is doing as she has been working on this for far longer than I have (just tinkering with it this morning)
Yes, mine as @vvcb if for the feature engineering part. I will work on this during the week.
Thanks,
On my side, will push the feature engineering code for grouping the chief complaint (https://github.com/LTHTR-DST/hdruk_avoidable_admissions/issues/36) tomorrow morning and let you know if I'm facing any problem. Apologies, got delayed with other projects
On my side, will push the feature engineering code for grouping the chief complaint (#36) tomorrow morning and let you know if I'm facing any problem. Apologies, got delayed with other projects
https://github.com/LTHTR-DST/hdruk_avoidable_admissions/pull/40
https://docs.google.com/spreadsheets/d/18XbVmWJsccACoTDFd8EBeslKPtsPApqi/edit#gid=589962562