Province estimates

This page gives the province estimates from the joint model, whose methods are on the Methods page. The national estimates are the sum of the provinces here. The per-province forecast is on the province forecasts page and its scoring is in the forecast by province.

This page is generated from docs/pages/estimates/province.jl. The model code it calls is in src/. See aim and origins and limitations.

Load packages, data and fitted chains

# Shared setup: packages, observations and the fit registry. See
# `docs/pages/_setup.jl`.
using BVDOutbreakSize
include(joinpath(pkgdir(BVDOutbreakSize), "docs", "pages", "_setup.jl"))

validation_forecast_from (generic function with 1 method)

# The fits this page reads, loaded from the cache here.
chn_joint = load_fit("joint");

Last updated: 23 October 2026.

Data as of: 20 October 2026.

Summary

Compute the weekly cases and reproduction number by province

# The provincial Rt trajectories, rebuilt on the same grid as the
# reproduction number by province figure further down.
province_summary_rt = reconstruct_patch_rt(
    chn_joint;
    n = obs.n, breakpoint = _BREAKPOINT, n_patches = N_PATCHES,
    rt_start = _rt_start_plot,
    rt_walk_start = clamp(_BREAKPOINT - RT_WALK_LEAD, _rt_start_plot, obs.n),
    ramp = RT_INTERVENTION_RAMP
);
# The week-ahead province forecast, read from the same draws as the
# province forecasts page.
province_summary_forecast = forecast_provinces(
    fit_forecast("joint"); horizon = 7, n_patches = N_PATCHES
);
province_cases_rt = cases_rt_table(
    province_summary_rt, province_cases.days, province_cases.increments;
    cutoff = obs.cutoff, n = obs.n, forecast = province_summary_forecast
);
province_cases_rt_fig = plot_cases_rt(province_cases_rt);

Each province's confirmed cases over the most recent week against its reproduction number, with the week before and the forecast week ahead. Faint bars span the 90% credible interval of the reproduction number, and the forecast also carries a bar across its 90% interval for cases.

Compute the province comparison

province_headline_md = patch_headline(chn_joint, N_PATCHES)
province_headline = Markdown.parse(province_headline_md);

Province	Share of infections (%)	R at the cut-off	P(R > 1)	CFR (%)	Relative ascertainment
Ituri	64–72	0.71–1.52	60%	36.0–60.7	0.89–1.23
Nord-Kivu	22–30	0.72–1.58	64%	39.1–63.8	0.76–1.08
Haut-Uele	4–5	0.6–1.67	52%	36.1–60.4	0.89–1.23
Sud-Kivu	0–0	0.49–1.57	42%	36.4–60.8	0.76–1.23
Tshopo	1–1	0.78–1.88	79%	36.4–61.0	0.85–1.25
Bas-Uele	0–0	0.63–1.71	60%	36.2–61.0	0.83–1.3
Sud-Ubangi	0–0	0.62–1.7	60%	36.4–61.0	0.76–1.25
Nord-Ubangi	0–0	0.82–2.72	86%	36.5–61.1	0.78–1.28

Ituri has the most infections with probability over 99%, and 7 of 8 provinces are more likely than not to be growing. Infections imported from another province make up 0.2–2.8% of infections to date.

The provinces compared from the joint posterior, as equal-tailed 90% credible intervals, with shares and probabilities computed draw by draw. The reproduction number and the relative case ascertainment are identified only as a product, and the per-province deaths break the tie.

Maps

Maps of infections, reproduction number and ascertainment

# The infections panel is also drawn alone for the summary page.
province_infections_panel = (;
    values = province_map_summary(chn_joint, :C_T_patch, N_PATCHES).values,
    title = "Infections to date", scale = log10,
    colorbar_label = "Infections (median)",
)
province_map_fig = plot_province_map(
    [
        province_infections_panel,
        (;
            province_map_summary(chn_joint, :R_T_patch, N_PATCHES)...,
            title = "Reproduction number at the cut-off",
            diverging_at = 1.0, colorbar_label = "R (median)",
        ),
        (;
            province_map_summary(
                chn_joint, :province_ascertainment, N_PATCHES
            )...,
            title = "Relative case ascertainment", diverging_at = 1.0,
            scale = log10, colorbar_label = "Relative to national (median)",
        ),
    ]
);

Each province is shaded by its posterior median. The reproduction number at the cut-off and the relative case ascertainment are centred on one, and a province whose 90% credible interval spans one is washed out.

Detail by province

Each province's case-fatality ratio is in the case-fatality ratio by province.

Compute the per-province tables and figure

# The tables below and the pair-plot dropdowns further down name the
# provinces in this order.
@assert PROVINCE_LABELS[1:N_PATCHES] == [
    "Ituri", "Nord-Kivu", "Haut-Uele",
    "Sud-Kivu", "Tshopo", "Bas-Uele", "Sud-Ubangi", "Nord-Ubangi",
]
province_detail_tables = [
    patch_summary_table(chn_joint, N_PATCHES; patch = p) for p in 1:N_PATCHES
];
province_detail_fig = plot_patch_summary(chn_joint, N_PATCHES);

Each province's estimates as medians with 30%, 60% and 90% credible intervals, repeated in the tables below. The log-Rt deviation is from the national trend, the drift is the deviation's walk scale and the contrast is against Ituri.

Ituri

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	10008.0	12086.0	13528.0	16272.0	18172.0	22365.0
2	Reproduction number	0.71	0.87	0.97	1.13	1.25	1.52
3	Daily infections at cut-off	42.0	59.0	71.0	98.0	117.0	166.0
4	log-Rt deviation from trend	-0.31	-0.18	-0.11	0.01	0.09	0.23
5	log-Rt vs primary patch	0.0	0.0	0.0	0.0	0.0	0.0
6	Rt deviation innovation sd	0.053	0.071	0.082	0.104	0.12	0.154
7	Relative case ascertainment	0.89	0.96	1.0	1.05	1.11	1.23

Nord-Kivu

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	3697.0	4510.0	5047.0	6217.0	6943.0	8606.0
2	Reproduction number	0.72	0.89	0.99	1.17	1.3	1.58
3	Daily infections at cut-off	33.0	47.0	57.0	79.0	94.0	137.0
4	log-Rt deviation from trend	-0.32	-0.15	-0.07	0.04	0.11	0.26
5	log-Rt vs primary patch	-0.38	-0.16	-0.05	0.12	0.23	0.42
6	Rt deviation innovation sd	0.047	0.066	0.078	0.105	0.126	0.16
7	Relative case ascertainment	0.76	0.85	0.9	0.98	1.01	1.08

Haut-Uele

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	627.0	769.0	872.0	1070.0	1205.0	1498.0
2	Reproduction number	0.6	0.78	0.9	1.15	1.31	1.67
3	Daily infections at cut-off	4.0	5.0	7.0	10.0	13.0	20.0
4	log-Rt deviation from trend	-0.51	-0.3	-0.17	0.03	0.14	0.35
5	log-Rt vs primary patch	-0.54	-0.29	-0.14	0.09	0.22	0.49
6	Rt deviation innovation sd	0.101	0.123	0.137	0.165	0.183	0.22
7	Relative case ascertainment	0.89	0.97	1.0	1.06	1.1	1.23

Sud-Kivu

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	2.0	4.0	5.0	8.0	10.0	16.0
2	Reproduction number	0.49	0.69	0.82	1.06	1.23	1.57
3	Daily infections at cut-off	0.0	0.0	0.0	0.0	0.0	0.0
4	log-Rt deviation from trend	-0.75	-0.42	-0.26	-0.05	0.06	0.28
5	log-Rt vs primary patch	-0.75	-0.39	-0.23	0.01	0.13	0.38
6	Rt deviation innovation sd	0.068	0.094	0.111	0.141	0.164	0.203
7	Relative case ascertainment	0.76	0.91	0.97	1.02	1.07	1.23

Tshopo

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	129.0	163.0	186.0	238.0	272.0	353.0
2	Reproduction number	0.78	0.99	1.1	1.34	1.49	1.88
3	Daily infections at cut-off	2.0	3.0	4.0	6.0	7.0	11.0
4	log-Rt deviation from trend	-0.23	-0.06	0.03	0.17	0.26	0.44
5	log-Rt vs primary patch	-0.33	-0.08	0.05	0.25	0.38	0.62
6	Rt deviation innovation sd	0.059	0.077	0.09	0.113	0.131	0.162
7	Relative case ascertainment	0.85	0.95	0.99	1.04	1.09	1.25

Bas-Uele

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	16.0	22.0	26.0	35.0	42.0	57.0
2	Reproduction number	0.63	0.84	0.97	1.19	1.34	1.71
3	Daily infections at cut-off	0.0	0.0	0.0	1.0	1.0	1.0
4	log-Rt deviation from trend	-0.45	-0.21	-0.1	0.06	0.14	0.33
5	log-Rt vs primary patch	-0.53	-0.23	-0.09	0.13	0.26	0.52
6	Rt deviation innovation sd	0.057	0.078	0.092	0.12	0.138	0.177
7	Relative case ascertainment	0.83	0.94	0.98	1.04	1.1	1.3

Sud-Ubangi

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	2.0	3.0	5.0	7.0	10.0	15.0
2	Reproduction number	0.62	0.83	0.96	1.19	1.35	1.7
3	Daily infections at cut-off	0.0	0.0	0.0	0.0	0.0	0.0
4	log-Rt deviation from trend	-0.46	-0.22	-0.11	0.06	0.15	0.34
5	log-Rt vs primary patch	-0.55	-0.23	-0.09	0.13	0.26	0.53
6	Rt deviation innovation sd	0.055	0.075	0.089	0.119	0.138	0.174
7	Relative case ascertainment	0.76	0.92	0.97	1.02	1.08	1.25

Nord-Ubangi

7×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Cumulative infections	4.0	6.0	8.0	14.0	19.0	34.0
2	Reproduction number	0.82	1.09	1.25	1.65	1.99	2.72
3	Daily infections at cut-off	0.0	0.0	0.0	0.0	1.0	2.0
4	log-Rt deviation from trend	-0.16	0.03	0.15	0.37	0.53	0.85
5	log-Rt vs primary patch	-0.24	0.03	0.18	0.44	0.63	1.0
6	Rt deviation innovation sd	0.066	0.093	0.109	0.139	0.161	0.206
7	Relative case ascertainment	0.78	0.92	0.97	1.02	1.09	1.28

Isolation beds

Province bed table

province_bed_overview = province_bed_table(chn_joint, N_PATCHES);

8×6 DataFrame

Row	Province	Beds	Bed demand	Occupied beds	Utilisation (%)	Shortfall
	String	String	String	String	String	String
1	Ituri	1074 (705–1553)	647 (576–722)	645 (569–720)	60.3 (40.7–93.6)	0 (0–0)
2	Nord-Kivu	302 (196–439)	399 (347–450)	302 (196–403)	100.0 (88.7–100.0)	96 (0–217)
3	Haut-Uele	147 (95–212)	72 (53–92)	71 (53–92)	49.1 (29.9–80.6)	0 (0–0)
4	Sud-Kivu	28 (17–43)	24 (17–34)	23 (16–30)	88.3 (50.8–100.0)	0 (0–10)
5	Tshopo	31 (20–46)	22 (16–28)	21 (16–28)	69.1 (41.7–100.0)	0 (0–3)
6	Bas-Uele	5 (2–10)	13 (9–18)	5 (2–10)	100.0 (100.0–100.0)	8 (2–14)
7	Sud-Ubangi	9 (4–17)	1 (0–2)	1 (0–2)	9.8 (3.0–31.3)	0 (0–0)
8	Nord-Ubangi	1 (0–9)	1 (0–3)	1 (0–2)	100.0 (13.2–100.0)	0 (0–1)

Beds, bed demand, occupied beds, utilisation and shortfall by province at the cut-off, from the occupancy and bed counts in the situation reports. Occupied beds are the modelled demand capped at the beds, and the shortfall is the demand above them.

Size and infections

Modelled infections by province

province_infections_fig = plot_infections_patches(
    chn_joint;
    n = obs.n, seeding = obs.seeding, n_patches = N_PATCHES
);

Modelled infections by province, daily (top) and cumulative (bottom).

The provinces are coupled by the gravity kernel in the mixing and importation Methods section. The distances between the patches' population centres run from 191 km between Sud-Ubangi and Nord-Ubangi to 1 220 km between Sud-Kivu and Sud-Ubangi. Ituri is 324 km from Nord-Kivu and 216 km from Haut-Uele. About half of what leaves Nord-Kivu lands in Sud-Kivu, 260 km away and the second most populous of the eight. Nord-Ubangi sends about a third of its exports to Sud-Ubangi, its nearest neighbour.

Importation intensity and imports by province

importation_table = summary_table(
    chn_joint, [:importation_epsilon];
    digits = 4,
    labels = Dict(:importation_epsilon => "Importation intensity")
);

province_imports_fig = plot_imports_patches(
    chn_joint;
    n = obs.n, seeding = obs.seeding, n_patches = N_PATCHES
);

1×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Importation intensity	0.0025	0.0061	0.0088	0.0142	0.0183	0.0265

Imported infections by province. Every arrival is debited from its origin on the same day, so the figure shows where infection occurred rather than extra infection.

Point prevalence

Point prevalence is the number of people infected at the cut-off who have not yet died, recovered or been detected as a suspected case. We split it into people infected but not yet symptomatic and people with symptoms who have not yet been detected. Together they are the infections in the community. The point prevalence section of the Methods gives the calculation. The table gives each rate per 100,000 residents and the number of people in the community, as medians with 90% credible intervals.

Compute point prevalence by province

province_prevalence = prevalence_overview(
    patch_prevalence(chn_joint, N_PATCHES),
    PROVINCE_POPULATIONS[1:N_PATCHES];
    labels = PROVINCE_LABELS[1:N_PATCHES], patch = collect(1:N_PATCHES)
);
province_prevalence_fig = plot_prevalence_ranking(
    province_prevalence; title = "Point prevalence by province"
);

Province	Population	Infected but not yet symptomatic per 100,000	Symptomatic, undetected per 100,000	In the community per 100,000	Number in the community
Ituri	4008000	13.1 (7.0–24.1)	15.1 (5.4–32.9)	28.4 (13.3–55.7)	1140 (532–2231)
Nord-Kivu	7574000	5.5 (2.9–10.4)	6.7 (2.7–13.9)	12.2 (6.0–23.4)	923 (454–1773)
Haut-Uele	2046000	2.6 (1.2–5.7)	3.1 (0.9–7.0)	5.8 (2.4–12.4)	118 (49–253)
Tshopo	2582000	1.1 (0.5–2.4)	1.1 (0.3–2.6)	2.1 (0.9–5.0)	55 (23–128)
Bas-Uele	1250000	0.2 (0.1–0.6)	0.2 (0.1–0.7)	0.5 (0.2–1.3)	6 (2–17)
Nord-Ubangi	1425000	0.1 (0.0–0.8)	0.1 (0.0–0.6)	0.2 (0.1–1.4)	3 (1–19)
Sud-Ubangi	2755000	0.0 (0.0–0.1)	0.0 (0.0–0.1)	0.0 (0.0–0.2)	1 (0–4)
Sud-Kivu	6565000	0.0 (0.0–0.0)	0.0 (0.0–0.0)	0.0 (0.0–0.1)	1 (0–3)

A province rate is an average over zones with different levels of transmission. The health-zone estimates give the same rates zone by zone.

Reproduction number by province

The national reproduction number is in reproduction number over time.

Reproduction number by province

province_rt_fig = plot_rt_patches(
    chn_joint;
    n = obs.n, breakpoint = _BREAKPOINT,
    n_patches = N_PATCHES,
    rt_start = _rt_start_plot,
    rt_walk_start = clamp(_BREAKPOINT - RT_WALK_LEAD, _rt_start_plot, obs.n),
    display_start = _rt_start_plot,
    as_of_date = string(obs.cutoff), seeding = obs.seeding,
    ramp = RT_INTERVENTION_RAMP
);

The reproduction number of each province, with the national trajectory in grey. The deviations sum to zero, so the grey band is the incidence-weighted middle of the panels rather than any one province. Sud-Kivu, Tshopo, Bas-Uele, Sud-Ubangi and Nord-Ubangi hold few confirmed cases, so their panels are carried mostly by the deviation prior and their widths are not measurements.

The spread of the provinces' log-Rt deviations is the spatial diagnostic. Its prior gives a 26% probability that the Ituri to Nord-Kivu ratio moves by more than 25% over the window. With eight patches and five of them carrying little signal, the cross-province correlation is not identified and tracks its prior.

Spatial hyperparameter summary table

# Labels for the spatial hyperparameters, shared by this table and the pair
# plot against their prior below.
spatial_labels = Dict(
    :region_sd => "Rt deviation spread",
    :region_halflife => "Rt deviation half-life (days)",
    :region_corr_primary_secondary => "Ituri-N.Kivu Rt correlation",
    :province_ascertainment_sd => "Ascertainment spread",
    :importation_epsilon => "Importation intensity",
    :importation_destination_sd => "Importation destination spread",
    :importation_origin_destination_correlation =>
        "Origin-destination correlation",
    :importation_flow_sd => "Importation flow spread",
    :importation_reciprocity => "Importation reciprocity",
    :province_cfr_sd => "Lethality spread (logit)",
    :province_death_ascertainment_sd => "Death-confirmation spread"
)
spatial_hyper_table = summary_table(
    chn_joint,
    [
        :region_sd, :region_halflife, :region_corr_primary_secondary,
        :province_ascertainment_sd,
    ];
    digits = 3, labels = spatial_labels
);

4×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Rt deviation spread	0.007	0.033	0.065	0.137	0.187	0.294
2	Rt deviation half-life (days)	25.716	36.649	44.297	63.739	80.475	120.579
3	Ituri-N.Kivu Rt correlation	-0.727	-0.489	-0.309	0.028	0.226	0.523
4	Ascertainment spread	0.014	0.047	0.078	0.154	0.204	0.338

The spatial hyperparameters of the province deviations. The correlation is between deviation innovations that sum to zero, so it leans negative, and with equal spreads each province's correlations with the other seven average −1/7.

Case-fatality ratio by province

The case-fatality ratio of each province is set against the national ratios in the confirmed case-fatality ratio. Only the product of a province's lethality and death confirmation is identified, and the province compositions Methods section says how the priors split it.

Province case-fatality spread

province_cfr_spread = summary_table(
    chn_joint,
    [:province_cfr_sd, :province_death_ascertainment_sd];
    digits = 3, labels = spatial_labels
);

2×7 DataFrame

Row	Quantity	Lower 90%	Lower 60%	Lower 30%	Upper 30%	Upper 60%	Upper 90%
	String	Float64	Float64	Float64	Float64	Float64	Float64
1	Lethality spread (logit)	0.011	0.046	0.078	0.15	0.2	0.294
2	Death-confirmation spread	0.008	0.03	0.052	0.1	0.133	0.189

Province case-fatality table

confirmed_cfr = delay_corrected_confirmed_cfr(
    chn_joint;
    obs_confirmed = obs.confirmed_cases,
    obs_confirmed_deaths = obs.confirmed_deaths
);
province_cfr = province_cfr_table(
    chn_joint, confirmed_cfr;
    province_cases = vec(sum(province_cases.increments; dims = 2)),
    province_deaths = vec(sum(province_deaths.increments; dims = 2)),
    n_patches = N_PATCHES
);

8×4 DataFrame

Row	Province	Naive observed confirmed ratio	Delay-corrected confirmed CFR	Structural (infection-based) CFR
	String	String	String	String
1	Ituri	46.5%	43.6% (36.5–52.1%)	48.1% (36.0–60.7%)
2	Nord-Kivu	55.9%	56.4% (47.4–67.9%)	50.9% (39.1–63.8%)
3	Haut-Uele	43.6%	43.9% (36.2–52.8%)	48.5% (36.1–60.4%)
4	Sud-Kivu	33.3%	46.9% (34.3–65.3%)	48.9% (36.4–60.8%)
5	Tshopo	35.5%	45.6% (34.9–58.3%)	48.7% (36.4–61.0%)
6	Bas-Uele	30.8%	45.6% (32.8–61.2%)	48.8% (36.2–61.0%)
7	Sud-Ubangi	50.0%	47.4% (34.5–65.2%)	49.2% (36.4–61.0%)
8	Nord-Ubangi	33.3%	46.7% (33.2–63.8%)	48.8% (36.5–61.1%)

Province parameters against their priors

The pair plots below set the province parameters against their prior, drawn from the province model.

Draw from the patch model's prior

prior_patch_chn = patch_prior_draws(obs);

Spatial hyperparameter pair plot (prior overlaid)

spatial_pair_fig = plot_pair(
    chn_joint,
    [
        :region_sd, :region_halflife, :region_corr_primary_secondary,
        :province_ascertainment_sd, :importation_epsilon,
        :importation_destination_sd,
        :importation_origin_destination_correlation, :importation_flow_sd,
        :importation_reciprocity, :province_cfr_sd,
        :province_death_ascertainment_sd,
    ];
    prior = prior_patch_chn, labels = spatial_labels
);

Joint posterior of the spatial hyperparameters, with the prior overlaid. These are the spread, half-life and correlation of the Rt deviations, the spread of case ascertainment, the importation intensity, its destination and flow spreads and their correlations, and the spreads of lethality and death confirmation.

Each province's pair plot sets the reproduction number at the cut-off against the relative case ascertainment, the case-fatality ratio against the relative death confirmation, and the importation intensity against the export weight relative to Ituri. Each pair is identified only as a product, so a ridge between the two is expected and the prior sets the position along it.

Compute the per-province pair plots

province_pair_labels = Dict(
    :R_T_patch => "Reproduction number",
    :province_ascertainment => "Case ascertainment",
    :CFR_patch => "Case-fatality ratio",
    :province_death_ascertainment => "Death confirmation",
    :importation_epsilon_patch => "Importation intensity",
    :export_weight => "Export weight"
)
# Ituri is the export reference, its weight fixed at one, so its pair plot
# leaves the weight out.
province_pair_figs = [
    plot_pair(
        chn_joint,
        [
            :R_T_patch, :province_ascertainment,
            :CFR_patch, :province_death_ascertainment,
            :importation_epsilon_patch,
            (p == 1 ? () : (:export_weight,))...,
        ];
        patch = p, prior = prior_patch_chn, labels = province_pair_labels
    )
        for p in 1:N_PATCHES
];

Ituri pair plot (prior overlaid)

Nord-Kivu pair plot (prior overlaid)

Haut-Uele pair plot (prior overlaid)

Sud-Kivu pair plot (prior overlaid)

Tshopo pair plot (prior overlaid)

Bas-Uele pair plot (prior overlaid)

Sud-Ubangi pair plot (prior overlaid)

Nord-Ubangi pair plot (prior overlaid)

Composition checks

Whether the model reproduces each province's observed share of the national total is on the in-sample checks page.

Data currency

The province blocks are the provincial tables of the situation reports: confirmed cases, confirmed deaths and the 24-hour laboratory volumes. A block that stops updating shows its last date before the cut-off rather than a flat series.

Province block currency

# Every province block whose last vintage falls before the cut-off, from
# the shared stream registry rather than a per-page list of dates. The
# grace the national table allows is not applied here: one vintage behind
# is already worth reading.
province_currency = let
    status = stream_report_status(obs; stratum = :province)
    behind = status[[ismissing(d) || d > 0 for d in status.days_since], :]
    isempty(behind) ?
        Markdown.parse(
            "Every province block reports to the cut-off, $(obs.cutoff)."
        ) :
        MarkdownTable(
            DataFrame(
                "Block" => behind.label,
                "Last reported" => [
                    ismissing(d) ? "never" : string(d)
                    for d in behind.last_date
                ],
                "Days before cut-off" => [
                    ismissing(d) ? "-" : string(d)
                    for d in behind.days_since
                ]
            )
        )
end;

Every province block reports to the cut-off, 2026-10-20.

Province fit diagnostics

The national fit diagnostics summarise the same joint fit as a whole, and its breakdown by parameter covers every parameter.

Province-level parameter diagnostics

# The quantities the sections above hold one per province: the
# reproduction number and infections to date from the renewal equation,
# the case-fatality ratio and the two ascertainment ratios the province
# compositions identify, and the importation intensity and export weight
# the per-province pair plots show.
province_diagnostic_quantities = [
    :R_T_patch, :C_T_patch, :CFR_patch, :province_ascertainment,
    :province_death_ascertainment, :importation_epsilon_patch, :export_weight,
]
province_diagnostics_labels = merge(
    province_pair_labels, Dict(:C_T_patch => "Infections to date")
)
province_diagnostics_frame = let df = parameter_diagnostics(chn_joint)
    df[in.(Symbol.(df.parameter), Ref(province_diagnostic_quantities)), :]
end
province_diagnostics_table = MarkdownTable(
    family_diagnostics_table(
        province_diagnostics_frame;
        n = length(province_diagnostic_quantities),
        labels = province_diagnostics_labels
    )
);

parameter	elements	max_rhat	min_ess_bulk	above_1.1
Importation intensity	8	1.006	282	0
Infections to date	8	1.006	554	0
Reproduction number	8	1.006	606	0
Case-fatality ratio	8	1.009	666	0
Death confirmation	8	1.003	887	0
Case ascertainment	8	1.003	958	0
Export weight	7	1.001	1586	0

Sampler diagnostics for the province-level parameters of the joint fit: the reproduction number, infections to date, case-fatality ratio, case and death ascertainment ratios, importation intensity and export weight.

Saving province assets

The summary dashboard shows the province comparison, the province map, the reproduction number by province and the weekly cases against the reproduction number, so they are written here.

Write the province dashboard assets

dashboard_dir = joinpath(
    pkgdir(BVDOutbreakSize), "docs", "src", "summary_assets"
)
mkpath(dashboard_dir)
CairoMakie.save(joinpath(dashboard_dir, "rt_provinces.png"), province_rt_fig)
# The summary's cases against R figure is drawn at the page width with
# larger labels.
CairoMakie.save(
    joinpath(dashboard_dir, "cases_rt_provinces.png"),
    plot_cases_rt(province_cases_rt; size = (680, 560), fontsize = 13)
)
# The summary's province map: infections to date alone.
CairoMakie.save(
    joinpath(dashboard_dir, "province_map.png"),
    plot_province_map(
        [province_infections_panel]; ncols = 1, panel_size = (680, 500)
    )
)
open(joinpath(dashboard_dir, "provinces.md"), "w") do io
    print(io, province_headline_md)
end
# The summary's compact province table: infections to date, R at the
# cut-off and the confirmed cases over the latest week of reports, the
# same week the cases against R figure ends on.
province_recent_cases = let last_day = min(last(province_cases.days), obs.n)
    weekly_count_table(
        province_cases.days, province_cases.increments[1:N_PATCHES, :],
        PROVINCE_LABELS[1:N_PATCHES];
        cutoff = obs.cutoff - Day(obs.n - last_day), n = last_day, weeks = 1
    ).count
end
open(joinpath(dashboard_dir, "provinces_summary.md"), "w") do io
    print(
        io,
        province_summary_table(
            chn_joint, N_PATCHES; recent_cases = province_recent_cases
        )
    )
end

---

The full analysis code, data and model definitions are in the epiforecasts/BVDOutbreakSize repository.