PSM

PSM

data=read.csv(file.choose(), header=T) # Import data called data
View(data)# View data
duplicated(data$Dup) #Check for duplicates on column Dup, not a very efficient way for large dataset
data$Dup[duplicated(data1$Dup)] #Show duplicate entries, don't like how results are displayed
unique(data1[duplicated(data1$Dup),])#Better way of showing unique repeat entries

#PSM
setwd("C:/users/machariam/Desktop/Rwanda PW Impact report")
getwd()
RwandaData=read.csv("CleanV4.csv")
attach(RwandaData)
RwandaData=as.data.frame(na.omit(RwandaData)) #Omit missing values

#sapply(RwandaData, function(x) sum(is.na(x)))#Inspect missing values
#RwandaData=read.csv("CleanV5.csv", header=T, na.strings=c(""," ","NA")) #Insert NAs on missing values
na.omit(RwandaData)
#detach(RwandaData)
attach(RwandaData)
RwandaData[1:10,]
RwandaData[75:79,]
install.packages("MatchIt")
library(MatchIt)
names(RwandaData)
str(RwandaData)

#Nearest Neighbor

m.out = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                data = RwandaData, method = "nearest",
                ratio = 1) 
summary(m.out)
plot(m.out, type = "jitter")
plot(m.out, type = "hist")

#Nearest Neighbor with caliper

m.out1 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                data = RwandaData, method = "nearest", caliper=0.16)
plot(m.out1, type = "hist")
summary(m.out1)
plot(m.out1, type = "jitter") 

#Exact Matching

m.out2 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                 data = RwandaData, method = "exact")
summary(m.out2)
plot(m.out2, type = "hist")

#Subclassification

m.out3 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                data = RwandaData, method = "subclass", subclass=4) 
summary(m.out3)
plot(m.out3, type = "hist")


#Optimal Matching

install.packages("optmatch")
library(optmatch)
m.out4 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                 data = RwandaData, method = "optimal", ratio=1) 

#Coarsened Exact Matching

install.packages("cem")
library(cem)
m.out5 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                 data = RwandaData, method = "cem")
summary(m.out5)
plot(m.out5, type = "hist")

m.out6 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                 data = RwandaData, method = "genetic")
plot(m.out6, type = "hist")


m.out7 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio+TLU,
                data = RwandaData, method = "nearest",
                distance ="logit") 
plot(m.out7, type = "hist")

m.out8 = matchit(Use ~ FarmingImportanceCode + TotalHHMembers  + NumberBtw14To65 + RespSexCode +RespAge+RespEducCode+TotalMaizeLand +YoungDepRatio,
                 data = RwandaData, method = "nearest", caliper=0.16)
plot(m.out8,type = "hist")
summary(m.out8)



m.data1 <- match.data(m.out)
write.csv(m.data1, file = ("C:/users/machariam/Desktop/Reduced.csv"))