Skip to main content

Table 3 Posterior estimates obtained from the geostatistical zero-inflated negative binomial (ZINB) models

From: Bayesian variable selection in modelling geographical heterogeneity in malaria transmission from sparse data: an application to Nouna Health and Demographic Surveillance System (HDSS) data, Burkina Faso

Parameters

An. funestus

An. gambiae

Median (95% BCI)

Median (95% BCI)

Intercept

-0.38 (-1.09, 0.46)

1.85 (-0.11, 3.56)

Year2

0.24 (-0.64, 1.07)

-1.01 (-2.75, 0.39)

Rainfall

0.87 (0.021, 1.71)

-2.33 (-4.67, -0.2)

Vegetation (NDVI)

1.12 (0.63, 1.65)

-

Day temp (LSTD)

-0.78 (-1.56, 0.00)

-

Night temp (LSTN)

-

-1.3 (-1.9, -0.64)

Amplitude

3.53 (3.50, 3.56)

5.88 (5.83, 5.93)

Shift/phase

0.648 (0.645, 0.653)

-1.159 (-1.163, -1.155)

Dispersion (r)

0.45 (0.32, 0.63)

0.93 (0.71, 1.24)

Spatial variation

0.90 (0.33, 2.03)

0.54 (0.22, 1.13)

Range (km)a

1 (3, 83)

5 (1, 35)

Temporal variation

-

0.76 (0.48, 1.26)

Parameters

Mixing proportion

An. funestus

An. gambiae

Median (95% BCI)

Median (95% BCI)

Intercept

-14.42 (-27.08, -3.57)

-11.06 (-18.80, -3.47)

Year2

0.24 (-0.64, 1.07)

4.42 (-5.67, 11.66)

Distance to water body

-3.52 (-12.43, 3.51)

-

Rainfall

-

-

Vegetation (NDVI)

-

-

Day temp (LSTD)

4.14 (-9.38, 21.23)

3.21 (0.89, 5.89)

Night temp (LSTN)

5.13 (-5.08, 14.83)

-1.99 (-5.57, -0.57)

Amplitude

21.92 (21.69, 22.16)

4.03 (3.96, 4.18)

Shift/phase

1.06 (1.05, 1.07)

2.25 (2.22, 2.27)

  1. aminimum distance in kilometer at which the spatial correlation remains important, BCI = Bayesian Credible Interval.