Closed gvegayon closed 3 weeks ago
Attention: Patch coverage is 97.01493%
with 2 lines
in your changes missing coverage. Please review.
Project coverage is 94.29%. Comparing base (
394a03d
) to head (1af7ba5
). Report is 1 commits behind head on main.
Files | Patch % | Lines |
---|---|---|
...odel/src/pyrenew/model/rtinfectionsrenewalmodel.py | 87.50% | 1 Missing :warning: |
model/src/pyrenew/process/rtperiodicdiff.py | 97.87% | 1 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Working on the conflicts
Ready for review, @dylanhmorris and @damonbayer .
Looks good, but I had one question about
duration
. Why did we opt for this instead ofn_timepoints
? They appear to represent the same concept, but maybe I am missing something.
I am a bit fuzzy about this. The duration
keyword is used for processes only, while n_timepoints
for models. I don't have strong feelings about it, but duration
seemed like a natural name for processes, so I didn't want to change that. But the number of observations in a model didn't sound like duration
, so I opted for using n_timepoints
. Perhaps this could be homogenized to length
or something? I suggest continuing this discussion in an issue.
Happy to standardize on n_timepoints
. Probably a separate PR.
This PR adds the following:
process
module:RtPeriodicDiffProcess
andRtWeeklyDiffProcess(RtPeriodicDiffProcess)
(an instance).RandomVariable
samples from an autoregressive process of the following form: $\log Rt(t) \sim Normal(\log Rt(t - 1) + \beta (\log Rt(t-1) - \log Rt(t-2)), \sigma^2_r)$RandomVariable
s.