obi-eneanya/Mapping-the-pre-control-prevalence-of-LF-in-Nigeria

# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # # ~~~~~~~~ Modelling LF prevalence in Nigeria using a machine learning approach ~~~~~~~~~~~~~ # # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # # Date: 07.2018 # Place: London, UK # Project: Modelling LF prevalence in Nigeria using Quantile Regression Forest # Section 1: Set up a function to install and load multiple R packages. # Check to see if packages are installed. Install them if they are not, then load them into the R session. ipak Pred ICT Obs.Mf.Pred.Mf Obs Mf # 5.3 - Correlation tests cor.test(ICT.cor$Prevalence, Obs.ICT.Pred.ICT) #correlation between Obs ICT and Pred ICT cor.test(Mf.cor$Prevalence, Obs.Mf.Pred.Mf) #correlation between Obs Mf and Pred Mf # Section 6 - Function for plotting graphs of observed vs predicted values and 95% prediction intervals ggvalidate = low & observed Pred ICT # Now re-run the above commands to extract back the values. obs.ICT <- Obs.Pred.ICT[,2] # Predicted lower and upper bounds pred025 <- Predict.ICT.output[,3] pred975 <- Predict.ICT.output[,5] ICT.CI <- as.data.frame(cbind(pred025, pred975)) install.packages(c("ggplot2", "ggpubr", "ggsci")) library(ggplot2) source("ggvalidate.R") ggvalidate(observed = obs.ICT, predicted = pred.ICT, CI = cbind(pred025, pred975), type = 1, alpha = 1) ggvalidate(observed = obs.ICT, predicted = pred.ICT, CI = cbind(pred025, pred975), type = 2, alpha = 1) # Save workspace # END OF RUN