library(tidyverse)
library(scales)
library(janitor)
library(paletteer)
dir.create("Results", showWarnings = FALSE)
source("Scripts/Load_IMPLAN.r")
SINGLE <- FALSE

OUTPUT_RES <- function(SINGLE=FALSE){
	IMPLAN <- GET_IMPLAN_DATA(SINGLE)


IMPLAN_DOLLAR <- IMPLAN %>% filter(Impact!='Employment')

IMPLAN_DOLLAR <-  IMPLAN_DOLLAR  %>% group_by(Type,Year,Impact,Source) %>% mutate(value=ifelse(Region=='US',value-min(value),value) ) %>% ungroup
TEMP <- IMPLAN_DOLLAR %>% filter(Year==2028)  %>% mutate(value=0)
IMPLAN_DOLLAR <- rbind(IMPLAN_DOLLAR,rbind(TEMP %>% mutate(Year=2025),TEMP %>% mutate(Year=2026),TEMP %>% mutate(Year=2027)))
IMPLAN_DOLLAR$Impact <- factor(IMPLAN_DOLLAR$Impact,levels=c('Output','Value Added','Income','Federal','State','County'))
IMPLAN_DOLLAR$Region <- factor(ifelse(IMPLAN_DOLLAR$Region=="WY",'Wyoming',"Rest of United States"),levels=rev(c("Wyoming","Rest of United States")))
IMPLAN_DOLLAR_REGION <- IMPLAN_DOLLAR %>% filter(Impact %in% c('Income','Output','Value Added')) %>% group_by(Year,Region,Impact) %>% summarize(value=sum(value)) %>% ungroup

########################Plot of total impact in US and in Wyoming
IMPLAN_REGION_DOLLAR_PLOT <- ggplot(IMPLAN_DOLLAR_REGION ,aes(x=Year,y=value/10^9,fill=Region,group=Region))+geom_area()+facet_wrap(~Impact,ncol=1)+ylab("Billion Dollars")+theme_bw()+ theme(plot.title = element_text(size = 22, face = "bold"), axis.title = element_text(size = 18),axis.text  = element_text(size = 16),legend.text = element_text(size = 14),strip.text = element_text(size = 14),legend.position="top")+  theme(legend.title=element_blank())+scale_fill_manual(values=c("#3C3B6E","#FFC425"))+scale_x_continuous(breaks=seq(2025,2060,by=5))+scale_y_continuous(breaks=seq(0,5000,by=1))
########################Taxes
IMPLAN_TAX <- IMPLAN_DOLLAR %>% filter(Impact %in% c("County","State","Federal"))
OLD <- IMPLAN_TAX 
IMPLAN_TAX <- IMPLAN_TAX %>% group_by(Type,Year,Impact,Source) %>% mutate(value=ifelse(Region=='Rest of United States',value-min(value),value) ) %>% ungroup

IMPLAN_TAX <- IMPLAN_TAX %>% mutate(Impact=ifelse(Region=='Rest of United States' & Impact!='Federal',"Other States",as.character(Impact)))
IMPLAN_TAX$Impact <- ifelse(IMPLAN_TAX$Impact =='State',"Wyoming State Taxes",IMPLAN_TAX$Impact)
IMPLAN_TAX$Impact <- ifelse(IMPLAN_TAX$Impact =='County',"Wyoming County Taxes",IMPLAN_TAX$Impact)
IMPLAN_TAX$Impact <- factor(IMPLAN_TAX$Impact,levels=c('Output','Value Added','Income','Federal','Other States','Wyoming State Taxes','Wyoming County Taxes'))

#IMPLAN_TAX <- IMPLAN_TAX[!(IMPLAN_TAX$Type=='Direct' & IMPLAN_TAX$Region=='Rest of United States'& IMPLAN_TAX$Impact!='Federal'),] 
#IMPLAN_TAX <-IMPLAN_TAX[!(IMPLAN_TAX$Type=='Direct' & IMPLAN_TAX$Region=='Wyoming'& IMPLAN_TAX$Impact=='Federal'),] 
IMPLAN_DOLLAR <- IMPLAN_DOLLAR %>% filter(!(Impact %in% c("County","State","Federal","Other States"))) %>% rbind(IMPLAN_TAX)
IMPLAN_TAX_FIG_DATA <- IMPLAN_TAX  %>% group_by(Year,Impact) %>% summarize(value=sum(value)) %>% ungroup

TAX_PLOT <- ggplot(IMPLAN_TAX_FIG_DATA, aes(x=Year,y=value/10^6,fill=Impact,group=Impact))+geom_area()+ylab("Million Dollars")+theme_bw()+ theme(plot.title = element_text(size = 22, face = "bold"), axis.title = element_text(size = 18),axis.text  = element_text(size = 16),legend.text = element_text(size = 14),strip.text = element_text(size = 14),legend.position="top")+  theme(legend.title=element_blank())+scale_fill_manual(values=c("#3C3B6E","firebrick","#FFC425","#492F24"))+scale_x_continuous(breaks=seq(2025,2060,by=5))+scale_y_continuous(breaks=seq(0,5000,by=50))

#############Dollar values table
NPV <- IMPLAN_DOLLAR %>% mutate(value=value/10^9,Discount_0=value,Discount_2=value/(1+0.02)^(Year-2028),Discount_5=value/(1+0.05)^(Year-2028),Discount_10=value/(1+0.1)^(Year-2028)) %>% group_by(Impact,Region) %>% summarize(Discount_0=sum(Discount_0),Discount_2=sum(Discount_2),Discount_5=sum(Discount_5),Discount_10=sum(Discount_10)) %>% ungroup
NPV[,3:6] <-round(NPV[,3:6],2)
WY_NPV <- cbind(c("Discount","0%","2%","5%","10%"),NPV %>% filter(Region=='Wyoming') %>% select(-Region) %>% t) %>% as_tibble
colnames(WY_NPV ) <- WY_NPV[1,]
WY_NPV <- WY_NPV[-1,] 
WY_NPV <- WY_NPV %>%  mutate_if(is.character, str_trim)
WY_NPV[,-1] <- lapply(	lapply(WY_NPV[,-1] ,as.numeric) %>% as_tibble,dollar) %>% as_tibble
WY_TAX <- WY_NPV[,c(1,5:7)]
WY_NPV <- WY_NPV[,c(1:4)]


#US
US_NPV <- cbind(c("Discount","0%","2%","5%","10%"),NPV %>% filter(Region!='Wyoming') %>% select(-Region) %>% t) %>% as_tibble
colnames(US_NPV ) <- US_NPV[1,]
US_NPV <- US_NPV[-1,] 
US_NPV <- US_NPV %>%  mutate_if(is.character, str_trim)
US_NPV[,-1] <- lapply(	lapply(US_NPV[,-1] ,as.numeric) %>% as_tibble,dollar) %>% as_tibble
US_TAX <- US_NPV[,c(1,5:6)]
US_NPV <- US_NPV[,c(1:4)]
ALL_TAX <- US_TAX
ALL_TAX[,2] <- dollar(parse_number(as.character(US_TAX[,2] %>% t))+ parse_number(as.character(WY_TAX[,2] %>% t)))
ALL_TAX <- cbind(ALL_TAX,WY_TAX[,-1:-2])

#write.csv(US_TAX,"./Results/US_Tax_Values.csv",row.names=FALSE)



############Employment
IMPLAN_EMPLOY <- IMPLAN %>% filter(Impact=='Employment',Source=='IMPLAN') %>% group_by(Year,Region,Impact) %>% summarize(value=sum(value)) %>% ungroup

IMPLAN_EMPLOY <-  IMPLAN_EMPLOY  %>% group_by(Year,Impact) %>% mutate(value=ifelse(Region=='US',value-min(value),value) ) %>% ungroup
TEMP <- IMPLAN_EMPLOY %>% filter(Year==2028)  %>% mutate(value=0)
IMPLAN_EMPLOY <- rbind(IMPLAN_EMPLOY,rbind(TEMP %>% mutate(Year=2025),TEMP %>% mutate(Year=2026),TEMP %>% mutate(Year=2027)))


IMPLAN_EMPLOY$Region <- factor(ifelse(IMPLAN_EMPLOY$Region=="WY",'Wyoming',"Rest of United States"),levels=rev(c("Wyoming","Rest of United States")))
 IMPLAN_EMPLOYMENT_PLOT <- ggplot(IMPLAN_EMPLOY ,aes(x=Year,y=value,fill=Region,group=Region))+geom_area()+ylab("Employment (Full-Time Equivalent)")+theme_bw()+ theme(plot.title = element_text(size = 22, face = "bold"), axis.title = element_text(size = 18),axis.text  = element_text(size = 16),legend.text = element_text(size = 14),strip.text = element_text(size = 14),legend.position="top")+  theme(legend.title=element_blank())+scale_fill_manual(values=c("#3C3B6E","#FFC425"))+scale_x_continuous(breaks=seq(2025,2060,by=5))+scale_y_continuous(breaks=c(seq(0,50000,by=1000)),labels = label_comma())


############Save results
HEADER <- ifelse(SINGLE,"Single Phase: ","")

write.csv(WY_NPV,paste0("./Results/",HEADER,"Wyoming_NPV_Values.csv"),row.names=FALSE)
write.csv(US_NPV,paste0("./Results/",HEADER,"US_NPV_Values.csv"),row.names=FALSE)
write.csv(NPV,paste0("./Results/",HEADER,"NPV_Values.csv"),row.names=FALSE)
write.csv(ALL_TAX,paste0("./Results/",HEADER,"Tax_Values.csv"),row.names=FALSE)

ggsave( filename = paste0("./Results/",HEADER,"Dollar Values of Model.png"),  plot = IMPLAN_REGION_DOLLAR_PLOT,  width = 8.5,  height = 11*(3./4),  units = "in", dpi = 300)
ggsave( filename = paste0("./Results/",HEADER,"Tax Model.png"),  plot = TAX_PLOT,  width = 8.5,  height = 11*(3./4),  units = "in", dpi = 300)
ggsave( filename = paste0("./Results/",HEADER,"Employment.png"),  plot = IMPLAN_EMPLOYMENT_PLOT ,  width = 8.5,  height = 11*(2/4),  units = "in", dpi = 600)
}

OUTPUT_RES(TRUE)
OUTPUT_RES(FALSE)

#########
if(exists("RES")){rm(RES)}
	for(YEAR in 2028:2036){
TEMP  <-  read_csv(paste0("Model_Outputs/IMPLAN/Wyoming/WY_",YEAR,"/occupation_impacts_table.csv"))[,c(2:3,5)]
colnames(TEMP) <- c("Occupation","Employment","Wages")
TEMP$Wages <- parse_number(TEMP$Wages)
TEMP <- TEMP %>%  mutate(Year=YEAR)
if(!exists("RES")){RES <- TEMP}else{RES <- rbind(RES,TEMP)}
rm(TEMP)
	}

if(exists("RES_US")){rm(RES_US)}
	for(YEAR in 2028:2036){
TEMP  <-  read_csv(paste0("Model_Outputs/IMPLAN/US/US_",YEAR,"/occupation_impacts_table.csv"))[,c(2:3,5)]
colnames(TEMP) <- c("Occupation","Employment","Wages")
TEMP$Wages <- parse_number(TEMP$Wages)
TEMP <- TEMP %>%  mutate(Year=YEAR)
if(!exists("RES_US")){RES_US <- TEMP}else{RES_US <- rbind(RES_US,TEMP)}
rm(TEMP)
	}
JOBS <- rbind(RES %>% filter(Year<2036) %>% group_by(Occupation) %>% summarize(Peak_Employment=max(Employment),Employment=mean(Employment),Wages=mean(Wages)) %>% ungroup %>% arrange(desc(Employment)) %>% mutate(Region='Wyoming',Period='Construction'),
RES %>% filter(Year==2036) %>% arrange(desc(Employment)) %>% mutate(Peak_Employment=Employment,Region='Wyoming',Period='Operations')%>% select(-Year),
RES_US %>% filter(Year<2036) %>% group_by(Occupation) %>% summarize(Peak_Employment=max(Employment),Employment=mean(Employment),Wages=mean(Wages)) %>% ungroup %>% arrange(desc(Employment)) %>% mutate(Region='United States',Period='Construction'),
RES_US %>% filter(Year==2036) %>% arrange(desc(Employment)) %>% mutate(Peak_Employment=Employment,Region='United States',Period='Operations')%>% select(-Year))
###

WY_JOBS <- JOBS %>% filter(Region=='Wyoming') %>%  rename(Peak_Wy_Employment=Peak_Employment,Wy_Employment=Employment,Wy_Wages=Wages)  %>% select(-Region)
US_JOBS <- JOBS %>% filter(Region!='Wyoming') %>%  rename(Peak_US_Employment=Peak_Employment,US_Employment=Employment,US_Wages=Wages)  %>% select(-Region)
JOINED <- WY_JOBS %>% left_join(US_JOBS)  %>% mutate(US_Employment=US_Employment-Wy_Employment,Peak_US_Employment=Peak_Wy_Employment,US_Wages=US_Wages-Wy_Wages) 
JOINED$US_Employment  <-  ifelse(JOINED$US_Employment<0,0,JOINED$US_Employment)
JOINED$US_Wages <-  ifelse(JOINED$US_Employment==0,0,JOINED$US_Wages)
JOINED$Peak_US_Employment <-  ifelse(JOINED$US_Employment==0,0,JOINED$Peak_US_Employment)
EMP_SUMMARY <- JOINED %>% mutate(Total_Employment=US_Employment+Wy_Employment,Total_Wages=US_Wages+Wy_Wages) %>% group_by(Period) %>% mutate(Rank=rank(-rank(Total_Employment))) %>% arrange(Period,Rank)



TOTAL_JOBS <- JOBS %>% group_by(Occupation,Period) %>% summarize(Total_Employment=sum(Employment,na.rm=TRUE),Total_Wages=sum(Wages,na.rm=TRUE)) %>% ungroup %>% group_by(Period) %>% mutate(Rank=rank(-rank(Total_Employment))) %>% arrange(Period,Rank) 

JOBS %>% group_by(Occupation,Period) %>% summarize(Total_Employment=sum(Employment),Total_Wages=sum(Wages))