Health-zone forecast evaluation

How the health-zone split on the health-zone forecasts page has scored against what each zone went on to report. The patch totals are scored as in the national forecast evaluation and the zone split against two persistence rules of its own. The same scoring by province is on the province forecast evaluation page. The in-sample zone checks are on the health-zone in-sample checks page.

Load packages, data and fitted chains

julia

# 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)

julia

# The zone fit melded from the frozen joint, loaded from the cache here.
frozen_local = load_fit("local_frozen_validation");

Summary

The overall bullets come first, then a block per province, from the scores below. The log score of the split is higher when the model allocates the week's cases better than the persistence rules.

Split: the zone model allocates the week better than both persistence rules in 3 of 5 provinces scored, over 492 observed confirmed cases.

Ituri

Log score of the split: -65.0, against -225.73 for share persistence and -87.26 for naive persistence.
Total over 265 observed cases: bias 0.4, within the 90% interval: true.

Nord-Kivu

Log score of the split: -38.36, against -80.82 for share persistence and -38.46 for naive persistence.
Total over 183 observed cases: bias 0.59, within the 90% interval: true.

Haut-Uele

Log score of the split: -9.69, against -10.45 for share persistence and -7.67 for naive persistence.
Total over 28 observed cases: bias 0.1, within the 90% interval: true.

Tshopo

Log score of the split: -17.85, against -15.15 for share persistence and -11.69 for naive persistence.
Total over 13 observed cases: bias -0.22, within the 90% interval: true.

Nord-Ubangi

Log score of the split: 0.0, against 0.0 for share persistence and 0.0 for naive persistence.
Total over 3 observed cases: bias -0.99, within the 90% interval: false.

Across releases

36 of 63 zones scored beat the persistence baseline, from the forecast made on 2026-09-27 onwards.

Forecast by health zone

The one-week-ahead zone forecast comes from the health-zone model melded from the frozen fit, run a week past the frozen cut-off. The observed count is the change in each zone's cumulative confirmed cases between the frozen cut-off and the vintage a week later, clamped at zero. The week is scored only when the zone tables carry a vintage on its last day. The split of each patch's weekly total is scored by its multinomial log score at the forecast shares. It is set against share persistence, the cumulative zone shares at the cut-off, and naive persistence, the zone split of the last seven days, each with a pseudo-count of 0.5 per zone. The patch total is scored against a naive persistence total with negative-binomial noise, as in the forecast scoring Methods section.

Zone forecast against observed

julia

# `frozen_zone_stage_inputs` is defined in the shared setup, so the zone
# estimates page reads the same frozen inputs.
frozen_zone_inputs = frozen_zone_stage_inputs(; forecast = true);
# The week is scored when the current zone tables carry a vintage a week
# past the frozen cut-off. Otherwise the section shows a note in place of
# its outputs.
zone_validation_horizon = frozen_zone_inputs.model_data.forecast.horizon
zone_truth = zone_forecast_truth(
    obs, frozen_zone_inputs;
    made_date = frozen_local.o.cutoff, horizon = zone_validation_horizon
)
_zone_validation_missing = Markdown.parse(
    "_The zone tables carry no vintage a week after the frozen cut-off, so " *
        "the zone forecast is not scored in this build._"
)
if any(!ismissing, zone_truth)
    zone_validation_forecast = zone_forecast(
        frozen_local.chn, frozen_zone_inputs
    )
    zone_validation_table = zone_forecast_vs_truth(
        zone_validation_forecast, frozen_zone_inputs; truth = zone_truth
    )
    zone_validation_scores = zone_forecast_scores(
        zone_validation_forecast, frozen_zone_inputs; truth = zone_truth
    )
    zone_validation_fig = plot_zone_forecast(
        zone_validation_table;
        patch_labels = frozen_zone_inputs.patch_labels,
        xlabel = "New confirmed cases over $(zone_validation_horizon) days",
        title = "Zone forecast from $(frozen_local.o.cutoff) against observed"
    )
else
    # The frames stay frames, since the release assets below write them.
    zone_validation_table = DataFrame()
    zone_validation_scores = DataFrame()
    zone_validation_fig = _zone_validation_missing
end;
# The scores rounded for display. Share persistence forecasts the split
# alone, so its total columns are blank. No patch is scored when every
# observed total is zero or incomplete, which leaves an empty frame
# without the columns.
zone_validation_scores_table = isempty(zone_validation_scores) ?
    _zone_validation_missing :
    let s = zone_validation_scores
        fmt(x, d) = ismissing(x) || isnan(x) ? "" :
        string(round(x; digits = d))
        DataFrame(
            "Province" => s.patch, "Rule" => s.method,
            "Observed total" => s.observed,
            "Log score of the split" => [fmt(x, 2) for x in s.log_score],
            "CRPS of the total" => [fmt(x, 1) for x in s.crps],
            "Log CRPS" => [fmt(x, 2) for x in s.log_crps],
            "Dispersion" => [fmt(x, 1) for x in s.dispersion],
            "Bias" => [fmt(x, 2) for x in s.bias],
            "Total within 90%" => [
                ismissing(x) ? "" : string(x)
                for x in s.coverage_90
            ]
        )
end;
zone_validation_table_display = isempty(zone_validation_table) ?
    _zone_validation_missing :
    zone_validation_table;

The fifteen zones with the largest forecast medians against what each went on to report, with every zone in the fold below the scores.

Province	Rule	Observed total	Log score of the split	CRPS of the total	Log CRPS	Dispersion	Bias	Total within 90%
Ituri	zone model	265	-65.0	28.7	0.1	20.3	0.4	true
Ituri	share persistence	265	-225.73
Ituri	naive persistence	265	-87.26	27.5	0.11	26.2	-0.14	true
Nord-Kivu	zone model	183	-38.36	37.5	0.17	18.8	0.59	true
Nord-Kivu	share persistence	183	-80.82
Nord-Kivu	naive persistence	183	-38.46	19.7	0.12	18.4	-0.16	true
Haut-Uele	zone model	28	-9.69	4.4	0.14	4.3	0.1	true
Haut-Uele	share persistence	28	-10.45
Haut-Uele	naive persistence	28	-7.67	3.2	0.12	3.2	-0.07	true
Tshopo	zone model	13	-17.85	2.1	0.17	1.8	-0.22	true
Tshopo	share persistence	13	-15.15
Tshopo	naive persistence	13	-11.69	1.7	0.14	1.5	-0.25	true
Nord-Ubangi	zone model	3	0.0	2.9	1.32	0.0	-0.99	false
Nord-Ubangi	share persistence	3	0.0
Nord-Ubangi	naive persistence	3	0.0	2.2	0.87	0.2	-0.98	false

Scores of each province's split and total for the model and the persistence rules. Share persistence forecasts the split alone, so its total columns are blank.

Zone forecast against observed, every zone

zone	patch	central_estimate	lower_90	lower_50	upper_50	upper_90	observed	within_90
Adja	Ituri	0	0	0	0	0	0	true
Ariwara	Ituri	0	0	0	0	1	0	true
Aru	Ituri	0	0	0	0	3	6	false
Aungba	Ituri	0	0	0	0	2	0	true
Bambu	Ituri	4	0	1	10	25	0	true
Boga	Ituri	0	0	0	0	4	0	true
Bunia	Ituri	71	28	50	97	147	55	true
Damas	Ituri	0	0	0	0	5	1	true
Drodro	Ituri	0	0	0	0	5	0	true
Fataki	Ituri	0	0	0	1	8	0	true
Gety	Ituri	0	0	0	0	7	0	true
Kambala	Ituri	0	0	0	0	1	0	true
Kilo	Ituri	0	0	0	0	6	0	true
Komanda	Ituri	8	0	3	17	35	6	true
Lita	Ituri	11	1	5	21	42	10	true
Logo	Ituri	0	0	0	0	4	0	true
Lolwa	Ituri	0	0	0	2	12	0	true
Mahagi	Ituri	0	0	0	0	1	0	true
Mambasa	Ituri	18	1	8	33	73	16	true
Mandima	Ituri	63	22	43	88	140	93	true
Mangala	Ituri	6	0	2	14	33	10	true
Mongbwalu	Ituri	4	0	1	11	27	4	true
Nia-Nia	Ituri	14	1	7	26	53	5	true
Nizi	Ituri	26	6	15	41	72	13	true
Nyankunde	Ituri	0	0	0	2	13	1	true
Rimba	Ituri	0	0	0	0	3	0	true
Rwampara	Ituri	24	6	14	38	65	39	true
Tchomia	Ituri	0	0	0	1	9	6	true
Patch total	Ituri	308	186	252	366	481	265	true
Alimbongo	Nord-Kivu	0	0	0	0	4	0	true
Beni	Nord-Kivu	101	49	76	133	193	71	true
Biena	Nord-Kivu	1	0	0	3	12	1	true
Butembo	Nord-Kivu	16	2	8	26	50	19	true
Goma	Nord-Kivu	0	0	0	0	0	0	true
Kalunguta	Nord-Kivu	14	2	7	24	42	8	true
Katwa	Nord-Kivu	26	7	16	39	64	16	true
Kayna	Nord-Kivu	0	0	0	0	2	0	true
Kyondo	Nord-Kivu	2	0	0	6	18	4	true
Lubero	Nord-Kivu	1	0	0	5	14	7	true
Mabalako	Nord-Kivu	18	3	10	30	53	17	true
Manguredjipa	Nord-Kivu	0	0	0	2	8	1	true
Masereka	Nord-Kivu	2	0	0	7	19	2	true
Musienene	Nord-Kivu	5	0	2	11	26	7	true
Mutwanga	Nord-Kivu	2	0	0	6	16	8	true
Oicha	Nord-Kivu	14	2	7	25	46	7	true
Vuhovi	Nord-Kivu	5	0	2	11	25	15	true
Patch total	Nord-Kivu	244	140	193	304	402	183	true
Boma Mangbetu	Haut-Uele	4	0	2	6	12	1	true
Dungu	Haut-Uele	0	0	0	0	1	0	true
Gombari	Haut-Uele	0	0	0	0	1	0	true
Isiro	Haut-Uele	2	0	0	4	8	4	true
Pawa	Haut-Uele	10	2	6	16	30	5	true
Rungu	Haut-Uele	0	0	0	0	2	1	true
Wamba	Haut-Uele	12	3	7	19	34	17	true
Patch total	Haut-Uele	30	8	19	43	75	28	true
Miti-Murhesa	Sud-Kivu	0	0	0	0	1	0	true
Patch total	Sud-Kivu	0	0	0	0	1	0	true
Bafwasende	Tshopo	1	0	0	2	4	8	false
Kabondo	Tshopo	1	0	0	2	4	3	true
Lubunga	Tshopo	0	0	0	1	2	0	true
Makiso-Kisangani	Tshopo	5	1	3	9	16	1	true
Mangobo	Tshopo	2	0	1	4	8	0	true
Tshopo	Tshopo	0	0	0	0	2	1	true
Wanie-Rukula	Tshopo	0	0	0	0	2	0	true
Patch total	Tshopo	11	2	7	17	30	13	true
Buta	Bas-Uele	0	0	0	1	2	0	true
Ganga	Bas-Uele	1	0	0	2	5	0	true
Viadana	Bas-Uele	1	0	0	1	4	0	true
Patch total	Bas-Uele	2	0	1	4	8	0	true
Bulu	Sud-Ubangi	0	0	0	1	2	0	true
Patch total	Sud-Ubangi	0	0	0	1	2	0	true
Wapinda	Nord-Ubangi	0	0	0	0	0	3	false
Patch total	Nord-Ubangi	0	0	0	0	0	3	false

Forecast by health zone across releases

The archived zone forecast of each release is scored against what each zone went on to report. The scores are those of the province forecast evaluation, with each zone scored as its own stream against a persistence baseline built from its own history. A week is left out when the zone tables carry no vintage on its last day, when it holds a harmonisation-break day, or when its province moved unallocated cases into named zones during it. The figures show the ten zones with the most observed cases over the scored weeks, and the folds hold every zone.

Load and summarise the zone forecast scores

julia

# Written by `scripts/score_releases.jl` from each release's
# `zone_forecast.csv`. Missing files read as empty tables.
zone_release_scores_df = _release_data(
    joinpath("zone", "forecast_scores.csv"),
    (;
        release = String, made_date = Date, stream = String, horizon = Int,
        target_date = Date, fit = String, crps = Float64,
        log_crps = Float64, dispersion = Float64, overprediction = Float64,
        underprediction = Float64, coverage_50 = Float64,
        coverage_90 = Float64,
        bias = Float64, n_samples = Int,
        log_rel_to_baseline = Float64,
    )
)
zone_release_overlay_df = scored_overlay(
    _release_data(
        joinpath("zone", "forecast_overlay.csv"),
        (;
            release = String, made_date = Date, stream = String,
            horizon = Int, target_date = Date, fit = String,
            observed = Float64, median = Float64, lo30 = Float64,
            hi30 = Float64, lo60 = Float64, hi60 = Float64, lo90 = Float64,
            hi90 = Float64,
        )
    )
)
# Every scored zone, ordered by its observed cases over the scored weeks.
zone_release_keys = let o = zone_release_overlay_df[
        zone_release_overlay_df.fit .== JOINT_FIT, :,
    ]
    tot = Dict{String, Float64}()
    for r in eachrow(o)
        k = zone_score_key(r.stream)
        k === nothing || (tot[k] = get(tot, k, 0.0) + r.observed)
    end
    sort!(collect(keys(tot)); by = k -> (-tot[k], k))
end
zone_release_labels = let lab = Dict(
        zip(frozen_zone_inputs.zone_keys, frozen_zone_inputs.zone_labels)
    )
    [get(lab, k, k) for k in zone_release_keys]
end
zone_release_top = first(zone_release_keys, 10)
_zone_release(tbl, keys = zone_release_keys) = zone_score_rows(
    tbl, keys, zone_release_labels[indexin(keys, zone_release_keys)]
)
# Zones in the order of `zone_release_keys`, under a `zone` column.
_zone_release_display(tbl) = let t = drop_degenerate_fit_column(
        drop_individual_fit_columns(tbl)
    )
    t = t[sortperm(indexin(t.stream, zone_release_labels)), :]
    DataFrame([(c == "stream" ? "zone" : c) => t[!, c] for c in names(t)])
end
zone_release_by_horizon = forecast_score_by_horizon(
    _zone_release(zone_release_scores_df, zone_release_top)
)
zone_release_by_release = forecast_score_by_release(
    _zone_release(zone_release_scores_df, zone_release_top)
)
zone_release_overview_display = _zone_release_display(
    forecast_score_overview(_zone_release(zone_release_scores_df))
)
zone_release_by_release_display = _zone_release_display(
    forecast_score_by_release(_zone_release(zone_release_scores_df))
);

_zone_release_empty = "No zone forecast has been scored yet. Scores " *
    "appear from the first release that archives the zone forecast onwards.";
# A plain statement above the tables: when nothing is scored yet, and
# otherwise the first cut-off scored.
zone_release_note = Markdown.parse(
    size(zone_release_scores_df, 1) == 0 ? _zone_release_empty :
        string(
            "Scores run from the forecast made on ",
            minimum(zone_release_scores_df.made_date), "."
        )
);

Scores run from the forecast made on 2026-09-27.

julia

zone_release_crps_fig = plot_forecast_crps_by_horizon(
    zone_release_by_horizon;
    title = "CRPS decomposition, by health zone",
    empty_message = _zone_release_empty
);

The mean CRPS split into its width, overprediction and underprediction, one panel per zone.

julia

zone_release_by_cutoff_fig = plot_forecast_skill_by_cutoff(
    zone_release_by_release;
    title = "Relative skill against the baseline by release, by health zone",
    empty_message = _zone_release_empty
);

Relative skill against the baseline release by release, on a log scale with the reference line at one.

julia

zone_release_overlay_fig = plot_forecast_overlay(
    _zone_release(zone_release_overlay_df, zone_release_top);
    empty_message = _zone_release_empty
);

Each release's zone forecasts against what each zone went on to report, by the cut-off each was made from. Each forecast shows its median and 90% predictive interval beside the persistence baseline and the observed count.

Zone scores across releases, every zone

zone	n	crps	rel_to_baseline	log_crps	log_rel_to_baseline	dispersion	overprediction	underprediction	coverage_50	coverage_90	bias
Mandima	4	10.48	1.7	0.136	1.55	9.31	0.53	0.64	1	1	0.04
Bunia	4	12.15	0.98	0.21	0.85	6.49	0.98	4.68	0.75	1	-0.22
Beni	4	9.75	2.28	0.13	1.89	7.78	1.97	0	0.75	1	0.27
Katwa	4	11.35	0.73	0.408	0.73	3.91	1.8	5.63	0.25	0.75	-0.13
Rwampara	4	9.44	1.22	0.327	1.01	5.07	2.75	1.61	0.5	1	0.17
Nizi	4	5.46	0.94	0.342	0.92	3.03	0.13	2.3	0.5	1	-0.34
Butembo	4	6.75	0.84	0.392	0.77	3.06	0.73	2.96	0.75	1	-0.14
Mambasa	4	7.62	1.23	0.904	2.2	1.71	0	5.91	0.5	0.75	-0.65
Mabalako	4	4.74	1.17	0.318	1.13	2.79	0.98	0.96	0.5	1	0.01
Nia-Nia	4	5.77	0.87	0.337	0.86	3.71	1.97	0.09	0.75	1	0.35
Wamba	4	3.25	1.01	0.28	0.82	1.65	0.09	1.51	0.5	1	-0.22
Musienene	4	4.97	0.67	0.518	0.67	1.93	0.12	2.92	0.5	1	-0.32
Mangala	4	3.48	0.77	0.374	0.84	2.1	0.02	1.36	0.75	1	-0.23
Lita	4	2.5	1.18	0.292	1.04	2.08	0	0.42	1	1	-0.24
Oicha	4	2.91	0.96	0.415	1.3	1.77	0.04	1.1	0.5	1	-0.34
Pawa	4	2.63	1.26	0.238	0.98	2.03	0.6	0	0.75	1	0.31
Kalunguta	4	3.55	3.81	0.273	2.46	2.62	0.94	0	0.75	1	0.34
Komanda	4	3.86	0.85	0.466	0.91	2.25	0.83	0.79	0.5	1	0.05
Bambu	4	2.61	0.75	0.473	0.85	1.4	0.05	1.16	0.75	1	-0.27
Vuhovi	4	2.57	1.19	0.498	1.05	1.55	0.64	0.38	0.5	1	0.04
Mongbwalu	4	3.51	1	0.455	0.86	2.41	1.1	0	1	1	0.47
Kyondo	4	1.68	1.2	0.45	1.24	0.94	0	0.74	0.75	1	-0.35
Mutwanga	4	2.31	0.65	0.626	0.92	0.67	0.01	1.63	0.5	1	-0.43
Lubero	4	1.58	1.55	0.415	1.63	0.76	0	0.82	0.5	1	-0.34
Masereka	4	1.64	1.5	0.37	1.57	1.21	0.07	0.36	0.75	1	-0.12
Rimba	4	2.7	0.87	1.005	1.13	0.05	0	2.65	0	0.75	-0.82
Bafwasende	4	2.7	0.92	0.675	0.88	0.23	0	2.46	0.5	0.75	-0.26
Makiso-Kisangani	4	1.5	1.1	0.333	0.94	0.97	0.49	0.05	0.75	1	0.19
Boma Mangbetu	4	0.94	0.85	0.281	0.82	0.68	0.15	0.11	1	1	0.08
Nyankunde	4	1.31	1.42	0.578	1.76	0.37	0	0.94	0.5	1	-0.53
Mangobo	4	1.42	0.74	0.562	0.8	0.35	0.13	0.94	0.25	0.75	-0.36
Gety	4	1.5	1.87	0.805	1.95	0.06	0	1.44	0	1	-0.81
Isiro	4	1.28	1.02	0.484	0.95	0.54	0.23	0.51	0.5	1	0.03
Biena	4	1.17	0.57	0.476	0.56	0.55	0.06	0.56	0.5	1	-0.04
Kabondo	4	0.72	0.73	0.339	0.72	0.28	0	0.44	0.5	1	-0.33
Ganga	4	0.73	0.59	0.382	0.57	0.16	0	0.57	0.5	1	-0.23
Rungu	4	0.72	0.56	0.43	0.61	0.05	0	0.67	0.25	0.75	-0.54
Kilo	4	0.83	0.63	0.416	0.71	0.07	0	0.76	0.5	1	-0.29
Lolwa	4	0.65	1.19	0.365	1.6	0.31	0	0.34	0.75	1	-0.42
Aru	4	0.6	1.62	0.352	1.72	0.03	0	0.57	0.5	1	-0.34
Fataki	4	0.61	0.61	0.328	0.73	0.15	0	0.46	0.75	1	-0.2
Manguredjipa	4	0.56	0.76	0.264	0.66	0.43	0	0.13	1	1	0.05
Kayna	4	0.5	0.37	0.266	0.36	0.12	0	0.38	0.75	1	-0.02
Adja	4	0.23	0.39	0.16	0.39	0.01	0	0.22	0.75	0.75	-0.18
Logo	4	0.23	0.47	0.152	0.46	0.05	0	0.18	0.75	1	-0.07
Alimbongo	3	0.3	0.11	0.18	0.17	0.15	0	0.15	1	1	-0.01
Tshopo	4	0.21	0.33	0.144	0.32	0.08	0	0.13	0.75	1	0.04
Buta	4	0.08	24.74	0.052	22.16	0.08	0	0	1	1	0.26
Viadana	4	0.22	0.28	0.141	0.33	0.22	0	0	1	1	0.42
Dungu	4	0.01	2.83	0.007	2.69	0.01	0	0	1	1	0.1
Gombari	4	0.01	0.45	0.009	0.41	0.01	0	0	1	1	0.11
Ariwara	4	0.03	0.38	0.014	0.3	0.03	0	0	1	1	0.11
Aungba	4	0.03	0.66	0.014	0.46	0.03	0	0	1	1	0.12
Boga	4	0.02	0.73	0.009	0.46	0.02	0	0	1	1	0.09
Damas	4	0.1	0.35	0.048	0.29	0.1	0	0	1	1	0.2
Drodro	4	0.11	0.94	0.053	0.73	0.11	0	0	1	1	0.23
Kambala	4	0.02	1.27	0.01	0.88	0.02	0	0	1	1	0.1
Mahagi	4	0.03	1.6	0.016	1.01	0.03	0	0	1	1	0.12
Tchomia	4	0.31	0.76	0.132	0.61	0.31	0	0	1	1	0.33
Goma	4	0.03	9.45	0.014	6.74	0.03	0	0	1	1	0.12
Miti-Murhesa	4	0.15		0.093		0.15	0	0	1	1	0.34
Bulu	4	0.04	0.19	0.027	0.18	0.04	0	0	1	1	0.19
Lubunga	4	0.14	30.01	0.089	26.91	0.14	0	0	1	1	0.34
Wanie-Rukula	4	0.07	22.41	0.043	18.98	0.07	0	0	1	1	0.24

Zone scores by release, every zone

made_date	zone	n	crps	rel_to_baseline	log_crps	log_rel_to_baseline	dispersion	overprediction	underprediction	coverage_50	coverage_90	bias
2026-09-27	Mandima	1	9.99	3.13	0.146	2.76	8.5	1.49	0	1	1	0.27
2026-10-02	Mandima	1	10.68	0.88	0.156	0.91	8.13	0	2.54	1	1	-0.32
2026-10-06	Mandima	1	10.08	1.78	0.13	1.52	10	0.08	0	1	1	0.07
2026-10-10	Mandima	1	11.17	3.04	0.113	2.64	10.62	0.55	0	1	1	0.14
2026-09-27	Bunia	1	19.53	1.14	0.353	0.92	5.15	0	14.38	0	1	-0.78
2026-10-02	Bunia	1	10.92	0.82	0.189	0.81	6.98	3.94	0	1	1	0.44
2026-10-06	Bunia	1	7.76	0.83	0.116	0.67	7.65	0	0.11	1	1	-0.07
2026-10-10	Bunia	1	10.39	1.05	0.181	0.93	6.18	0	4.22	1	1	-0.45
2026-09-27	Beni	1	6.9	3.43	0.108	3.03	6.19	0.71	0	1	1	0.22
2026-10-02	Beni	1	9.33	1.73	0.129	1.36	7.67	1.66	0	1	1	0.26
2026-10-06	Beni	1	8.07	1.1	0.105	0.96	8.01	0.06	0	1	1	0.05
2026-10-10	Beni	1	14.69	6.22	0.178	5.06	9.23	5.46	0	0	1	0.53
2026-09-27	Katwa	1	18.76	1.02	0.7	0.89	2.62	0	16.14	0	0	-0.93
2026-10-02	Katwa	1	10.07	1.12	0.331	1.13	3.68	0	6.39	0	1	-0.63
2026-10-06	Katwa	1	6.03	0.4	0.192	0.44	4.64	1.39	0	1	1	0.34
2026-10-10	Katwa	1	10.53	0.55	0.408	0.58	4.72	5.81	0	0	1	0.72
2026-09-27	Rwampara	1	9.8	1.43	0.395	1.23	3.35	0	6.45	0	1	-0.64
2026-10-02	Rwampara	1	9.07	1.14	0.259	0.91	6.11	2.96	0	1	1	0.44
2026-10-06	Rwampara	1	12.94	1.18	0.48	1.08	5.14	7.8	0	0	1	0.76
2026-10-10	Rwampara	1	5.95	1.15	0.174	0.72	5.69	0.26	0	1	1	0.12
2026-09-27	Nizi	1	3.63	0.74	0.353	0.91	2.28	0	1.35	1	1	-0.4
2026-10-02	Nizi	1	6.5	1.34	0.438	1.34	2.79	0	3.71	0	1	-0.56
2026-10-06	Nizi	1	7.06	0.93	0.368	0.73	2.92	0	4.14	0	1	-0.61
2026-10-10	Nizi	1	4.64	0.8	0.208	0.78	4.13	0.51	0	1	1	0.22
2026-09-27	Butembo	1	11.15	0.78	0.774	0.7	1.52	0	9.63	0	1	-0.85
2026-10-02	Butembo	1	4.89	1.18	0.279	1.28	2.69	0	2.2	1	1	-0.51
2026-10-06	Butembo	1	5.09	0.91	0.246	0.84	3.78	1.31	0	1	1	0.4
2026-10-10	Butembo	1	5.85	0.73	0.271	0.67	4.25	1.6	0	1	1	0.43
2026-09-27	Mambasa	1	6.04	11.81	1.458	24.46	0.21	0	5.83	0	0	-0.92
2026-10-02	Mambasa	1	3.07	1.05	0.511	1.32	1.5	0	1.58	1	1	-0.48
2026-10-06	Mambasa	1	4.45	0.87	0.316	1.02	3.44	0	1.01	1	1	-0.33
2026-10-10	Mambasa	1	16.91	1.04	1.33	1.49	1.7	0	15.21	0	1	-0.89
2026-09-27	Mabalako	1	5.33	0.83	0.538	0.97	1.61	0	3.72	0	1	-0.69
2026-10-02	Mabalako	1	2.36	2.9	0.17	3.51	2.23	0	0.13	1	1	-0.15
2026-10-06	Mabalako	1	4.28	2.21	0.209	1.91	3.48	0.8	0	1	1	0.3
2026-10-10	Mabalako	1	6.97	1	0.354	0.85	3.83	3.14	0	0	1	0.59
2026-09-27	Nia-Nia	1	6.65	1.82	0.251	1.49	4.72	1.93	0	1	1	0.45
2026-10-02	Nia-Nia	1	8.24	0.56	0.434	0.58	4	4.24	0	0	1	0.65
2026-10-06	Nia-Nia	1	3.33	1.35	0.255	1.61	2.97	0	0.36	1	1	-0.19
2026-10-10	Nia-Nia	1	4.87	0.87	0.409	0.82	3.15	1.72	0	1	1	0.48
2026-09-27	Wamba	1	1.46	0.94	0.17	0.69	1.46	0	0	1	1	-0.01
2026-10-02	Wamba	1	1.95	1.17	0.21	0.82	1.58	0.38	0	1	1	0.33
2026-10-06	Wamba	1	4.56	0.77	0.383	0.63	1.49	0	3.07	0	1	-0.63
2026-10-10	Wamba	1	5.02	1.36	0.355	1.41	2.05	0	2.96	0	1	-0.56
2026-09-27	Musienene	1	8.91	0.74	0.972	0.63	1.06	0	7.85	0	1	-0.9
2026-10-02	Musienene	1	5.5	0.92	0.561	0.97	1.67	0	3.83	0	1	-0.65
2026-10-06	Musienene	1	3.19	0.49	0.289	0.52	2.71	0.48	0	1	1	0.29
2026-10-10	Musienene	1	2.26	0.45	0.25	0.62	2.26	0	0	1	1	-0.03
2026-09-27	Mangala	1	6.55	1.23	0.54	1.1	2.2	0	4.36	0	1	-0.61
2026-10-02	Mangala	1	2.16	0.41	0.278	0.65	2.11	0	0.04	1	1	-0.07
2026-10-06	Mangala	1	2.25	0.48	0.269	0.61	2.17	0.08	0	1	1	0.14
2026-10-10	Mangala	1	2.98	1.02	0.408	0.96	1.93	0	1.05	1	1	-0.38
2026-09-27	Lita	1	2.56	1.07	0.289	0.9	2.03	0	0.53	1	1	-0.32
2026-10-02	Lita	1	2.09	1.05	0.251	0.73	2.05	0	0.04	1	1	-0.08
2026-10-06	Lita	1	2.42	1.16	0.286	1.37	2.13	0	0.29	1	1	-0.23
2026-10-10	Lita	1	2.93	1.46	0.342	1.39	2.09	0	0.84	1	1	-0.35
2026-09-27	Oicha	1	1.97	0.34	0.541	0.82	0.76	0	1.21	0	1	-0.58
2026-10-02	Oicha	1	3.71	2.03	0.642	2.97	1.09	0	2.61	0	1	-0.65
2026-10-06	Oicha	1	2.76	0.7	0.257	0.75	2.18	0	0.58	1	1	-0.28
2026-10-10	Oicha	1	3.23	5.19	0.22	3.95	3.05	0.18	0	1	1	0.15
2026-09-27	Pawa	1	2.55	1.91	0.186	1.68	2.18	0.37	0	1	1	0.25
2026-10-02	Pawa	1	2.53	2.4	0.175	1.91	2.35	0.19	0	1	1	0.15
2026-10-06	Pawa	1	1.85	1.41	0.157	0.93	1.77	0.08	0	1	1	0.15
2026-10-10	Pawa	1	3.58	0.78	0.434	0.73	1.81	1.77	0	0	1	0.69
2026-09-27	Kalunguta	1	2.39	2.27	0.233	1.39	1.98	0.41	0	1	1	0.26
2026-10-02	Kalunguta	1	3.5	3.93	0.271	2.85	2.52	0.98	0	1	1	0.38
2026-10-06	Kalunguta	1	3.1	3.49	0.219	2.48	2.86	0.24	0	1	1	0.19
2026-10-10	Kalunguta	1	5.23	5.84	0.37	3.96	3.12	2.12	0	0	1	0.53
2026-09-27	Komanda	1	4.66	0.55	0.34	0.55	3.22	1.44	0	1	1	0.42
2026-10-02	Komanda	1	4.5	0.87	0.565	0.77	2.63	1.88	0	0	1	0.64
2026-10-06	Komanda	1	2.34	1.32	0.316	1.23	1.86	0	0.48	1	1	-0.27
2026-10-10	Komanda	1	3.95	1.39	0.642	1.47	1.29	0	2.66	0	1	-0.6
2026-09-27	Bambu	1	1.83	0.9	0.362	1.14	1.37	0	0.46	1	1	-0.34
2026-10-02	Bambu	1	4.83	1.19	0.81	1.09	1	0	3.83	0	1	-0.73
2026-10-06	Bambu	1	1.99	0.85	0.359	1.17	1.63	0	0.37	1	1	-0.27
2026-10-10	Bambu	1	1.79	0.32	0.359	0.42	1.58	0.21	0	1	1	0.28
2026-09-27	Vuhovi	1	2.14	3.39	0.577	4.65	0.68	0	1.46	0	1	-0.63
2026-10-02	Vuhovi	1	1.63	0.65	0.239	0.54	1.59	0	0.04	1	1	-0.09
2026-10-06	Vuhovi	1	1.98	2.71	0.268	1.78	1.92	0.06	0	1	1	0.1
2026-10-10	Vuhovi	1	4.54	0.95	0.907	0.77	2.02	2.52	0	0	1	0.78
2026-09-27	Mongbwalu	1	4.2	1.33	0.294	0.74	3.22	0.98	0	1	1	0.36
2026-10-02	Mongbwalu	1	3.21	0.89	0.376	0.77	2.33	0.88	0	1	1	0.45
2026-10-06	Mongbwalu	1	3.42	0.88	0.478	0.86	2.12	1.31	0	1	1	0.49
2026-10-10	Mongbwalu	1	3.23	0.95	0.673	1.01	1.99	1.24	0	1	1	0.58
2026-09-27	Kyondo	1	1.04	0.93	0.362	0.86	0.74	0	0.3	1	1	-0.34
2026-10-02	Kyondo	1	1.19	1.11	0.414	1.02	0.75	0	0.44	1	1	-0.36
2026-10-06	Kyondo	1	1.32	1.76	0.298	1.87	1.32	0	0	1	1	-0.05
2026-10-10	Kyondo	1	3.17	1.2	0.726	1.54	0.94	0	2.23	0	1	-0.66
2026-09-27	Mutwanga	1	2.28	3.04	0.81	4.46	0.32	0	1.96	0	1	-0.77
2026-10-02	Mutwanga	1	4.97	1.46	1.086	2.15	0.57	0	4.39	0	1	-0.84
2026-10-06	Mutwanga	1	1.06	0.27	0.315	0.47	0.89	0	0.18	1	1	-0.25
2026-10-10	Mutwanga	1	0.92	0.15	0.295	0.21	0.89	0.02	0	1	1	0.16
2026-09-27	Lubero	1	0.45	1.22	0.237	1.42	0.43	0	0.01	1	1	-0.21
2026-10-02	Lubero	1	0.63	1.81	0.255	1.51	0.63	0	0	1	1	-0.03
2026-10-06	Lubero	1	2.36	0.88	0.563	0.95	0.94	0	1.42	0	1	-0.56
2026-10-10	Lubero	1	2.87	4.34	0.604	6.38	1.03	0	1.84	0	1	-0.58
2026-09-27	Masereka	1	1.84	2.61	0.603	3.34	0.49	0	1.35	0	1	-0.68
2026-10-02	Masereka	1	1.29	1.1	0.261	0.96	1.21	0	0.08	1	1	-0.15
2026-10-06	Masereka	1	1.83	1.44	0.311	1.25	1.61	0.22	0	1	1	0.23
2026-10-10	Masereka	1	1.59	1.3	0.304	1.24	1.52	0.06	0	1	1	0.12
2026-09-27	Rimba	1	0.73	1.46	0.495	1.46	0.06	0	0.67	0	1	-0.72
2026-10-02	Rimba	1	5.37	0.99	1.682	1.05	0.02	0	5.34	0	0	-0.95
2026-10-06	Rimba	1	3.99	1.69	1.361	2.55	0.07	0	3.91	0	1	-0.9
2026-10-10	Rimba	1	0.71	0.17	0.483	0.45	0.06	0	0.66	0	1	-0.72
2026-09-27	Bafwasende	1	0.24	2.41	0.152	2.37	0.24	0	0	1	1	0.44
2026-10-02	Bafwasende	1	0.21	2.92	0.131	2.89	0.21	0	0	1	1	0.41
2026-10-06	Bafwasende	1	1.9	0.87	0.759	0.82	0.22	0	1.68	0	1	-0.91
2026-10-10	Bafwasende	1	8.43	0.9	1.657	0.82	0.26	0	8.17	0	0	-0.99
2026-09-27	Makiso-Kisangani	1	0.86	1.3	0.163	0.98	0.86	0	0	1	1	0.01
2026-10-02	Makiso-Kisangani	1	0.92	1.45	0.184	1.36	0.89	0.02	0	1	1	0.18
2026-10-06	Makiso-Kisangani	1	0.96	0.81	0.193	0.74	0.77	0	0.19	1	1	-0.28
2026-10-10	Makiso-Kisangani	1	3.27	1.09	0.791	0.94	1.35	1.92	0	0	1	0.85
2026-09-27	Boma Mangbetu	1	1.01	1.27	0.275	1.32	0.64	0	0.37	1	1	-0.43
2026-10-02	Boma Mangbetu	1	0.7	1.05	0.198	0.86	0.65	0	0.06	1	1	-0.18
2026-10-06	Boma Mangbetu	1	1.08	0.59	0.337	0.61	0.75	0.33	0	1	1	0.49
2026-10-10	Boma Mangbetu	1	0.97	0.85	0.312	0.83	0.69	0.29	0	1	1	0.46
2026-09-27	Nyankunde	1	1.24	1.94	0.64	2.25	0.15	0	1.09	0	1	-0.72
2026-10-02	Nyankunde	1	1.47	2.25	0.601	3.6	0.46	0	1.01	1	1	-0.53
2026-10-06	Nyankunde	1	1.81	1.48	0.768	1.61	0.23	0	1.58	0	1	-0.72
2026-10-10	Nyankunde	1	0.71	0.61	0.304	0.79	0.65	0	0.06	1	1	-0.17
2026-09-27	Mangobo	1	1.47	0.93	0.567	1.03	0.32	0	1.16	0	1	-0.76
2026-10-02	Mangobo	1	2.47	0.98	0.808	1.09	0.29	0	2.19	0	0	-0.92
2026-10-06	Mangobo	1	0.69	4.7	0.318	6.48	0.29	0	0.4	1	1	-0.53
2026-10-10	Mangobo	1	1.06	0.31	0.555	0.38	0.53	0.53	0	0	1	0.76
2026-09-27	Gety	1	1.53	1.05	0.818	1.08	0.05	0	1.48	0	1	-0.84
2026-10-02	Gety	1	1.53	1.02	0.827	1.04	0.05	0	1.48	0	1	-0.81
2026-10-06	Gety	1	1.41	9.72	0.749	13.53	0.09	0	1.32	0	1	-0.75
2026-10-10	Gety	1	1.54	13.79	0.825	21.54	0.05	0	1.5	0	1	-0.84
2026-09-27	Isiro	1	1.04	1.28	0.323	1.07	0.75	0.29	0	1	1	0.48
2026-10-02	Isiro	1	1.25	1.03	0.621	1.1	0.62	0.64	0	0	1	0.78
2026-10-06	Isiro	1	0.54	0.59	0.224	0.51	0.46	0	0.08	1	1	-0.27
2026-10-10	Isiro	1	2.3	1.11	0.768	1.05	0.35	0	1.94	0	1	-0.86
2026-09-27	Biena	1	1.01	0.99	0.505	0.98	0.28	0	0.74	0	1	-0.61
2026-10-02	Biena	1	2.05	1.06	0.686	1.14	0.53	0	1.52	0	1	-0.68
2026-10-06	Biena	1	0.92	0.36	0.411	0.37	0.71	0.22	0	1	1	0.61
2026-10-10	Biena	1	0.71	0.26	0.301	0.26	0.67	0.04	0	1	1	0.52
2026-09-27	Kabondo	1	0.26	6.34	0.156	5.34	0.26	0	0	1	1	0.43
2026-10-02	Kabondo	1	0.4	0.65	0.256	0.6	0.25	0	0.15	1	1	-0.33
2026-10-06	Kabondo	1	0.89	0.54	0.43	0.5	0.31	0	0.58	0	1	-0.7
2026-10-10	Kabondo	1	1.35	0.82	0.513	0.9	0.31	0	1.04	0	1	-0.73
2026-09-27	Ganga	1	0.13	4.41	0.087	4.23	0.13	0	0	1	1	0.34
2026-10-02	Ganga	1	1.26	0.78	0.641	0.75	0.12	0	1.14	0	1	-0.84
2026-10-06	Ganga	1	1.27	0.76	0.642	0.71	0.12	0	1.15	0	1	-0.88
2026-10-10	Ganga	1	0.25	0.15	0.158	0.17	0.25	0	0	1	1	0.45
2026-09-27	Rungu	1	1.57	0.84	0.83	0.82	0.04	0	1.54	0	0	-0.94
2026-10-02	Rungu	1	0.04	0.03	0.028	0.03	0.04	0	0	1	1	0.19
2026-10-06	Rungu	1	0.63	0.76	0.432	1.3	0.05	0	0.57	0	1	-0.72
2026-10-10	Rungu	1	0.62	0.89	0.429	0.89	0.05	0	0.57	0	1	-0.72
2026-09-27	Kilo	1	2.4	1.95	1.085	2.42	0.04	0	2.36	0	1	-0.88
2026-10-02	Kilo	1	0.71	0.62	0.487	1.21	0.05	0	0.66	0	1	-0.72
2026-10-06	Kilo	1	0.12	0.06	0.061	0.06	0.12	0	0	1	1	0.24
2026-10-10	Kilo	1	0.07	0.1	0.032	0.08	0.07	0	0	1	1	0.18
2026-09-27	Lolwa	1	0.66	0.87	0.329	1.18	0.44	0	0.21	1	1	-0.31
2026-10-02	Lolwa	1	0.62	1.31	0.352	1.6	0.3	0	0.32	1	1	-0.41
2026-10-06	Lolwa	1	0.63	1.39	0.341	1.68	0.35	0	0.28	1	1	-0.38
2026-10-10	Lolwa	1	0.68	1.38	0.439	2.07	0.13	0	0.55	0	1	-0.61
2026-09-27	Aru	1	0.03	0.77	0.015	0.57	0.03	0	0	1	1	0.13
2026-10-02	Aru	1	0.03	0.89	0.015	0.54	0.03	0	0	1	1	0.12
2026-10-06	Aru	1	0.76	1.13	0.517	1.13	0.04	0	0.72	0	1	-0.77
2026-10-10	Aru	1	1.6	2.13	0.86	2.78	0.04	0	1.56	0	1	-0.86
2026-09-27	Fataki	1	0.65	0.95	0.384	1.35	0.24	0	0.41	1	1	-0.49
2026-10-02	Fataki	1	1.48	1.61	0.788	1.79	0.06	0	1.41	0	1	-0.81
2026-10-06	Fataki	1	0.2	0.16	0.092	0.17	0.2	0	0	1	1	0.3
2026-10-10	Fataki	1	0.1	0.09	0.047	0.09	0.1	0	0	1	1	0.21
2026-09-27	Manguredjipa	1	0.26	0.21	0.139	0.19	0.26	0	0	1	1	0.38
2026-10-02	Manguredjipa	1	0.41	3.15	0.2	2.58	0.41	0	0	1	1	0.44
2026-10-06	Manguredjipa	1	0.91	0.69	0.434	0.64	0.42	0	0.49	1	1	-0.48
2026-10-10	Manguredjipa	1	0.65	3.16	0.281	2.68	0.63	0	0.02	1	1	-0.15
2026-09-27	Kayna	1	1.57	0.85	0.844	0.84	0.03	0	1.54	0	1	-0.87
2026-10-02	Kayna	1	0.1	0.06	0.047	0.05	0.1	0	0	1	1	0.21
2026-10-06	Kayna	1	0.16	0.09	0.083	0.09	0.16	0	0	1	1	0.29
2026-10-10	Kayna	1	0.18	10.98	0.089	8.55	0.18	0	0	1	1	0.3
2026-09-27	Adja	1	0	0.07	0.002	0.09	0	0	0	1	1	0.04
2026-10-02	Adja	1	0.91	1.09	0.626	1.1	0.01	0	0.9	0	0	-0.92
2026-10-06	Adja	1	0.02	0.03	0.009	0.02	0.02	0	0	1	1	0.09
2026-10-10	Adja	1	0.01	0.01	0.003	0.01	0.01	0	0	1	1	0.06
2026-09-27	Logo	1	0.05	0.48	0.026	0.34	0.05	0	0	1	1	0.16
2026-10-02	Logo	1	0.77	1.35	0.527	1.36	0.03	0	0.74	0	1	-0.79
2026-10-06	Logo	1	0.07	0.11	0.037	0.09	0.07	0	0	1	1	0.2
2026-10-10	Logo	1	0.04	0.06	0.019	0.05	0.04	0	0	1	1	0.15
2026-10-02	Alimbongo	1	0.59	0.21	0.384	0.45	0.13	0	0.46	1	1	-0.59
2026-10-06	Alimbongo	1	0.16	0.03	0.079	0.05	0.16	0	0	1	1	0.28
2026-10-10	Alimbongo	1	0.15	0.18	0.076	0.14	0.15	0	0	1	1	0.28
2026-09-27	Tshopo	1	0.05	8.28	0.03	7.67	0.05	0	0	1	1	0.2
2026-10-02	Tshopo	1	0.59	0.67	0.407	0.66	0.07	0	0.52	0	1	-0.67
2026-10-06	Tshopo	1	0.1	0.13	0.07	0.13	0.1	0	0	1	1	0.31
2026-10-10	Tshopo	1	0.1	0.12	0.068	0.11	0.1	0	0	1	1	0.3
2026-09-27	Buta	1	0.06	51.33	0.041	44.74	0.06	0	0	1	1	0.23
2026-10-02	Buta	1	0.06	14.74	0.042	13.27	0.06	0	0	1	1	0.24
2026-10-06	Buta	1	0.09	29.86	0.057	27.1	0.09	0	0	1	1	0.28
2026-10-10	Buta	1	0.11	23.72	0.068	21.29	0.11	0	0	1	1	0.3
2026-09-27	Viadana	1	0.27	0.11	0.168	0.14	0.27	0	0	1	1	0.46
2026-10-02	Viadana	1	0.2	0.31	0.125	0.3	0.2	0	0	1	1	0.4
2026-10-06	Viadana	1	0.17	2.76	0.107	2.59	0.17	0	0	1	1	0.37
2026-10-10	Viadana	1	0.27	4.9	0.162	4.44	0.27	0	0	1	1	0.44
2026-09-27	Dungu	1	0.01	1.7	0.006	1.65	0.01	0	0	1	1	0.09
2026-10-02	Dungu	1	0.01	3.32	0.009	3.21	0.01	0	0	1	1	0.11
2026-10-06	Dungu	1	0.01	3.41	0.006	3.28	0.01	0	0	1	1	0.09
2026-10-10	Dungu	1	0.01	3.41	0.008	3.13	0.01	0	0	1	1	0.1
2026-09-27	Gombari	1	0.02	0.4	0.011	0.37	0.02	0	0	1	1	0.12
2026-10-02	Gombari	1	0.01	0.32	0.006	0.29	0.01	0	0	1	1	0.09
2026-10-06	Gombari	1	0.02	0.74	0.013	0.68	0.02	0	0	1	1	0.13
2026-10-10	Gombari	1	0.01	0.39	0.008	0.35	0.01	0	0	1	1	0.1
2026-09-27	Ariwara	1	0.05	0.58	0.023	0.41	0.05	0	0	1	1	0.15
2026-10-02	Ariwara	1	0.03	0.37	0.013	0.31	0.03	0	0	1	1	0.12
2026-10-06	Ariwara	1	0.02	0.33	0.011	0.26	0.02	0	0	1	1	0.1
2026-10-10	Ariwara	1	0.02	0.21	0.008	0.18	0.02	0	0	1	1	0.09
2026-09-27	Aungba	1	0.03	0.83	0.014	0.54	0.03	0	0	1	1	0.12
2026-10-02	Aungba	1	0.02	0.5	0.01	0.37	0.02	0	0	1	1	0.1
2026-10-06	Aungba	1	0.04	1.04	0.02	0.76	0.04	0	0	1	1	0.14
2026-10-10	Aungba	1	0.03	0.41	0.012	0.29	0.03	0	0	1	1	0.11
2026-09-27	Boga	1	0.02	0.8	0.009	0.43	0.02	0	0	1	1	0.08
2026-10-02	Boga	1	0.03	1.79	0.013	1.14	0.03	0	0	1	1	0.11
2026-10-06	Boga	1	0.02	0.39	0.008	0.25	0.02	0	0	1	1	0.08
2026-10-10	Boga	1	0.01	0.56	0.006	0.41	0.01	0	0	1	1	0.08
2026-09-27	Damas	1	0.24	0.7	0.114	0.57	0.24	0	0	1	1	0.33
2026-10-02	Damas	1	0.05	0.19	0.026	0.16	0.05	0	0	1	1	0.17
2026-10-06	Damas	1	0.09	0.3	0.035	0.22	0.09	0	0	1	1	0.18
2026-10-10	Damas	1	0.04	0.14	0.017	0.11	0.04	0	0	1	1	0.13
2026-09-27	Drodro	1	0.11	0.75	0.051	0.53	0.11	0	0	1	1	0.22
2026-10-02	Drodro	1	0.09	0.8	0.045	0.66	0.09	0	0	1	1	0.22
2026-10-06	Drodro	1	0.14	1.42	0.069	1.13	0.14	0	0	1	1	0.27
2026-10-10	Drodro	1	0.09	0.91	0.047	0.73	0.09	0	0	1	1	0.23
2026-09-27	Kambala	1	0.01	1.15	0.007	0.84	0.01	0	0	1	1	0.09
2026-10-02	Kambala	1	0.02	1.21	0.007	0.75	0.02	0	0	1	1	0.09
2026-10-06	Kambala	1	0.03	2.07	0.016	1.42	0.03	0	0	1	1	0.12
2026-10-10	Kambala	1	0.02	0.8	0.009	0.57	0.02	0	0	1	1	0.09
2026-09-27	Mahagi	1	0.04	1.53	0.018	0.95	0.04	0	0	1	1	0.13
2026-10-02	Mahagi	1	0.02	1.05	0.01	0.65	0.02	0	0	1	1	0.1
2026-10-06	Mahagi	1	0.04	1.43	0.018	1	0.04	0	0	1	1	0.14
2026-10-10	Mahagi	1	0.04	3.07	0.016	1.82	0.04	0	0	1	1	0.12
2026-09-27	Tchomia	1	0.65	1.1	0.277	0.88	0.65	0	0	1	1	0.49
2026-10-02	Tchomia	1	0.3	0.92	0.124	0.7	0.3	0	0	1	1	0.32
2026-10-06	Tchomia	1	0.15	0.39	0.07	0.35	0.15	0	0	1	1	0.26
2026-10-10	Tchomia	1	0.13	0.4	0.058	0.33	0.13	0	0	1	1	0.24
2026-09-27	Goma	1	0.02	5.64	0.011	4.49	0.02	0	0	1	1	0.11
2026-10-02	Goma	1	0.02	4.25	0.01	3.51	0.02	0	0	1	1	0.11
2026-10-06	Goma	1	0.05	28.06	0.023	19.88	0.05	0	0	1	1	0.15
2026-10-10	Goma	1	0.03	10.4	0.013	6.35	0.03	0	0	1	1	0.11
2026-09-27	Miti-Murhesa	1	0.07		0.044		0.07	0	0	1	1	0.24
2026-10-02	Miti-Murhesa	1	0.15		0.101		0.15	0	0	1	1	0.37
2026-10-06	Miti-Murhesa	1	0.15		0.095		0.15	0	0	1	1	0.35
2026-10-10	Miti-Murhesa	1	0.21		0.132		0.21	0	0	1	1	0.4
2026-09-27	Bulu	1	0.04	0.05	0.028	0.05	0.04	0	0	1	1	0.19
2026-10-02	Bulu	1	0.04	2.61	0.025	2.56	0.04	0	0	1	1	0.19
2026-10-06	Bulu	1	0.03	1.1	0.017	1.08	0.03	0	0	1	1	0.15
2026-10-10	Bulu	1	0.06	3.18	0.037	3.02	0.06	0	0	1	1	0.22
2026-09-27	Lubunga	1	0.11	15.59	0.073	14.46	0.11	0	0	1	1	0.32
2026-10-02	Lubunga	1	0.11	29.12	0.074	26.88	0.11	0	0	1	1	0.32
2026-10-06	Lubunga	1	0.12	30.1	0.08	26.84	0.12	0	0	1	1	0.33
2026-10-10	Lubunga	1	0.2	60.94	0.13	52.14	0.2	0	0	1	1	0.41
2026-09-27	Wanie-Rukula	1	0.06	12.75	0.038	10.61	0.06	0	0	1	1	0.22
2026-10-02	Wanie-Rukula	1	0.07	24.34	0.044	21.51	0.07	0	0	1	1	0.25
2026-10-06	Wanie-Rukula	1	0.07	37.93	0.046	32.03	0.07	0	0	1	1	0.25
2026-10-10	Wanie-Rukula	1	0.07	26.06	0.046	21.76	0.07	0	0	1	1	0.24

Saving zone forecast outputs

The frozen zone fit's one-week-ahead forecast against the observed zone increments, and its scores against the two persistence rules, are written as release assets.

Write zone forecast outputs

julia

output_dir = get(
    ENV, "BVD_OUTPUT_DIR",
    joinpath(pkgdir(BVDOutbreakSize), "output")
)
mkpath(output_dir)
CSV.write(
    joinpath(output_dir, "zone_forecast_validation.csv"),
    zone_validation_table
)
CSV.write(
    joinpath(output_dir, "zone_forecast_scores.csv"),
    zone_validation_scores
)

"/home/runner/work/BVDOutbreakSize/BVDOutbreakSize/output/zone_forecast_scores.csv"

Write the summary bullets

julia

# The bullets under the summary heading at the top of the page. They read
# the scores built further down, so they are written here and read back
# when the site is assembled.
evaluation_forecast_zone_summary = let s = zone_validation_scores
    fmt(x) = ismissing(x) || (x isa Real && isnan(x)) ? "n/a" :
        string(round(x; digits = 2))
    if isempty(s)
        string(
            "- **Zone forecasts:** the zone tables end at or before the ",
            "frozen cut-off, so no zone forecast is scored in this build."
        )
    else
        patches = unique(s.patch)
        rule(p, m) = let r = s[(s.patch .== p) .& (s.method .== m), :]
            size(r, 1) == 0 ? nothing : first(eachrow(r))
        end
        beats(p) = let m = rule(p, "zone model")
            m === nothing ? false :
                all(
                    r -> r === nothing || m.log_score >= r.log_score,
                    [rule(p, "share persistence"), rule(p, "naive persistence")]
                )
        end
        observed_total = sum(s[s.method .== "zone model", :observed])
        overall = [
            string(
                "- **Split:** the zone model allocates the week better ",
                "than both persistence rules in ", count(beats, patches),
                " of ", length(patches), " provinces scored, over ",
                observed_total, " observed confirmed cases."
            ),
        ]
        function detail(p)
            m = rule(p, "zone model")
            m === nothing && return string("**", p, "**", "\n\n- Not scored.")
            return join(
                [
                    string("**", p, "**"), "",
                    string(
                        "- Log score of the split: ", fmt(m.log_score),
                        ", against ",
                        fmt(
                            rule(p, "share persistence") === nothing ?
                                missing : rule(p, "share persistence").log_score
                        ),
                        " for share persistence and ",
                        fmt(
                            rule(p, "naive persistence") === nothing ?
                                missing : rule(p, "naive persistence").log_score
                        ),
                        " for naive persistence."
                    ),
                    string(
                        "- Total over ", m.observed,
                        " observed cases: bias ", fmt(m.bias),
                        ", within the 90% interval: ",
                        ismissing(m.coverage_90) ? "n/a" :
                            string(m.coverage_90), "."
                    ),
                ], "\n"
            )
        end
        join(vcat([join(overall, "\n")], [detail(p) for p in patches]), "\n\n")
    end
end
dashboard_dir = joinpath(
    pkgdir(BVDOutbreakSize), "docs", "src", "summary_assets"
)
mkpath(dashboard_dir)
write(
    joinpath(dashboard_dir, "evaluation_forecast_zone.md"),
    evaluation_forecast_zone_summary
);

Add the across-release scores to the summary

julia

# A block on the zone forecasts scored across releases, after the
# per-province blocks above.
zone_release_summary = let ov = forecast_score_overview(
        _zone_release(zone_release_scores_df)
    )
    rows = filter(r -> !ismissing(r.rel_to_baseline), ov)
    size(rows, 1) == 0 ?
        "**Across releases**\n\n- $(_zone_release_empty)" :
        string(
            "**Across releases**\n\n- ", count(<(1), rows.rel_to_baseline),
            " of ", size(rows, 1), " zones scored beat the persistence ",
            "baseline, from the forecast made on ",
            minimum(zone_release_scores_df.made_date), " onwards."
        )
end
open(joinpath(dashboard_dir, "evaluation_forecast_zone.md"), "a") do io
    write(io, "\n\n", zone_release_summary)
end;

Health-zone forecast evaluation ​

Summary ​

Forecast by health zone ​

Forecast by health zone across releases ​

Saving zone forecast outputs ​

Health-zone forecast evaluation

Summary

Forecast by health zone

Forecast by health zone across releases

Saving zone forecast outputs