BDI 2016-20

[QLLZ]

Data pull: HASH: 6772ea9da4f43ddc3da11c494a573af2ca3dd3c8 config

[QLLZ]

Model run: HASH: 6772ea9da4f43ddc3da11c494a573af2ca3dd3c8

[QLLZ]

Country data report

[eclee25]

Convergence is mostly good, with some slight instability in rho, lambda, sd_w, and some of the w. There are large variations in od_param as well. Rhats look good. Modeled estimates at observation and grid level match the data well, however, and seem to correspond with the GAM input. Sampling and warmup times are short (~2 min each), but this is not necessarily a problem.

Opinion: Approve

[QLLZ]

Config

Data pull:

HASH: 211913e0eb7d570ed62c48bc1ae9ecd6671a6cbc

[QLLZ]

Failed stan model run log file

[QLLZ]

Modeled cases look good. Convergency looks good. Population comparison looks good.

One question is in Figure 18. In the southwest part of the map, for adm2, the max posterior probability is >100 , while in admin1, the max posterior is <1. will look into this.

[eclee25]

@QLLZ I'm a bit confused about how there are comments for this one when the stan model failed.

[QLLZ]

Rerun on dev_u_combs_fix

HASH: e56580fa5ddab00293e31ff90139351f80ae3f6c

[QLLZ]

Country data reports

[eclee25]

All diags look good. Approve

[QLLZ]

Data pull, model run:

HASH: 05f598b95b5d42ac0049963f8a0b7e04d6f1dbd7

[javierps]

Sep 2023 Production run: all OK.

Suggestion: accept.

[eclee25]

Approve