bokola / covariates Goto Github PK

This contains files that allows to reproduce the exploratory analysis, parameter estimation and spatial prediction for the geostatistical model containing all the coviariates presented in the paper "Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict."

The files are:

• "1 - Exploratory analysis.R" is the R script for: exploring relationships between covariates and prevalence; testing for residual spatial correlation.

• "2 - Parameter estimation.R" is the R script for carrying out the fitting of the geostatistical model using the Monte carlo maximum likelihood method.

• "3 - Spatial prediction.R" is the R sripts for carrying out the spatial prediction of prevalence over a regular grid.

• "TZ_2015.csv" is the file containing the data. Each row corresponds to a sampled community with total number of tested people ("Ex"), nuber of positive cases ("Pf") and geographical locations. The column "utm_x" is the x-coordinate and "utm_y" is the y-coordinate in UTM; longitude and latitude are also reported in "Long" and "Lat", respectively. The name of the covariates in the data correspond to the same names and abbreviation used in the main manuscript.

• "TZ_2015_fit.RData" is an R object containing the fitted geostatistical model that is loaded within the R code of "3 - Spatial prediction.R".

• "pred_objects.RData" is an R object containing the prediction locations and the values of the covariates at those locations, that are used in "3 - Spatial prediction.R".