#####
#####
#####read in appropriate libraries
library(AICcmodavg)
library(lme4)
library(lattice)
library(mgcv)
library(gamm4)
library(car)
library(cvTools)
library(influence.ME)
library(sp)
library(rgdal)
library(raster)
library(maptools)
library(ape)
library(geosphere)
library(geoR)
library(gstat)

#read in data
Ponded <- read.csv("BinomialOnlyRowcropsExtremeChangePresent3.csv", header = TRUE)
head(Ponded)
nrow(Ponded)

#####
#####
#####set nominal variables as factors
Ponded$Wetland = factor(Ponded$Wetland)
Ponded$Wettype = factor(Ponded$Wettype)
Ponded$Year = factor(Ponded$Year)
Ponded$agtype = factor(Ponded$agtype)

#####
#####
#####make predictions for ponded area manually
#with original data
Ponded$Pred.Log.Meters.ManualFixedOnly <- NA
for(i in 1:nrow(Ponded)){
	Ponded$Pred.Log.Meters.ManualFixedOnly[[i]] = 8.04202+
	+ 0.97839 * Ponded$SemiPerm[[i]]+
	- 0.90915 * Ponded$Temporary[[i]]+
	+ 0.66729 * Ponded$Dryland[[i]]+
	- 0.20840 * Ponded$Gravity[[i]]+
	- ((0.18190 * Ponded$log.NewfootPA[[i]] - 0.18190 * -4.23480) / 0.53077)+
	+ ((0.07812 * Ponded$espro2PostChange[[i]] - 0.07812 * 0.28529) / 0.47472)+
	+ ((0.30557 * Ponded$log.lsprsPostChange[[i]] - 0.30557 * 5.47634) / 0.29421)+
	+ ((0.41977 * Ponded$log.wprsPostChange[[i]] - 0.41977 * 4.35447) / 0.74091)+
	+ ((0.70207 * Ponded$wtmaxsfPostChange[[i]] - 0.70207 * 26.24143) / 4.34563)
}

#####
#####
#####calculate the inverse log for predicted values
Ponded$Pred.Meters.ManualFixedOnly <- exp(Ponded$Pred.Log.Meters.ManualFixedOnly)

#obtain mean flooded area
mean(Ponded$Pred.Meters.ManualFixedOnly)

#obtain total flooded area
sum(Ponded$Pred.Meters.ManualFixedOnly)