PACU/electives changes - Githubissues

hqva commented 1 year ago

This is clearly far from complete - but I think I'm at the point where a review would be helpful to make sure I’m on the right track.

I'm also clinical Wed-Fri week so perhaps a good time to checkpoint?

I have moved mock data from json files to a mock.db file. This was initially generated with sdv, but sdv copies text fields in a way that caused problems going forward. It does relational things reasonably but hand-editing was simpler.
The live table has columns for examples to demonstrate that data is pulled from the right parts of the mock database - surgical/ preassessment / echo / labs / obs
I will need to discuss what to show and how to show it without overwhelming. table is already too wide really for normal monitors! It may be that there should be further information hidden behind a click view or a pop up card.
I have documented bugs and issues at the moment #TODO comments to avoid cluttering the “issues” section with things only relevant to my branch
How should I manage code duplication between my HyMind branch and this branch and keeping things up to date?
I am aware the wrangle.py file is a bit unwieldy. In all honesty a complete rethink may be needed re current approach of small SQL pulls then python merging - may be faster to do more at the time of the SQL pulls?
- For the moment I have not changed this. Some of the wrangling is done with updated functions and some are using original functions written by jen to wrangle to features.
- Another issue with this is that getting things ready for the model transforms / summarises / discards information that clinicians may want to access directly - eg. Echo data - this factors in to the issue above re displaying data.
I have added Pete Odor's PACU dashboard to the nav menu and as a HTML iframe.
- I've put the dashboard file in assets/odor_dashboard.html. It is 9mb so fails the check_added_large_file hook and also has (aggregated but) real patient data