CodeForPhilly / CodeForPhilly/chime
["model"] Results from running files (models, __init__, parameters, constants, base, validators) from develop branch in notebook differ from CHIME web app census csv download
- Dominant language
- Python
- Stars
- 210
- Forks
- 153
- PR merge metrics
- No merged PRs in 30d
Description
### Summary
I work for a data science team at a health system, we have limited knowledge/prior experience running docker containers etc so we downloaded (via urllib.request) the files in the src/pnn_chime folder to run locally and output csvs for patient census in each one of our regions our health system covers. code is in attached notebook. fyi, was having problems getting the relative imports to run so I regexed them out so the file could be run by exec().
The output from doing this direct file download and running is different than the output from the CHIME web app even though the same files should be running in both examples
### Additional details
ex row of region data:
Pop | current_hosp | date first hosp | doubling_time | initial_infections | hosp_rate | icu_rate | vent_rate | hosp_los | icu_los | vent_los | market_share
-- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | --
604227 | 14 | 3-18-2020 | 7 | 61 | 0.025 | 0.01 | 0.0075 | 7 | 9 | 10 | 0.233
output for that region
[2020-04-05_projected_census.xlsx](https://github.com/CodeForPhilly/chime/files/4435710/2020-04-05_projected_census.xlsx)
python notebook file as text file:
[automated_notebook_to_text.txt](https://github.com/CodeForPhilly/chime/files/4435713/automated_notebook_to_text.txt)
### Suggested fix
unknown
Contributor guide
Research direction
Start with the attached automated_notebook_to_text.txt and the files under src/pnn_chime, including models, __init__, parameters, constants, base, and validators. Reproduce the local execution and compare its projected census CSV or XLSX with the CHIME web app download for the provided region; done means the discrepancy is explained and a suggested fix is identified.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100