####DUALITY###
##Load the Data

my.data <- read.table(pipe("pbpaste"), sep="\t", header = TRUE)

#Load MetaSEM package
library(metaSEM)

head(my.data)
colnames(my.data)

# locate studies with information on at least 1 correlation, and no missing sample size
keepstudy <- rowSums(is.na(my.data[,4:39]))!=36 & is.na(my.data$NUMBER.OF.BANKS)==FALSE


# keep only the studies with information
data <- my.data[keepstudy,]

# check data
head(data)
length(data)
summary(data)

# varnames and labels
nvar <- 9
varnames <- c("DUA","BNKRISK","BSIZE","LEV","AGE","LIQ","OWNCON","INST","PERF")
labels <- list(varnames,varnames)


# create list with correlation matrices
cordat <- list()
for (i in 1:nrow(data)){	
  cordat[[i]] <- vec2symMat(as.matrix(data[i,4:39]),diag = FALSE)
  dimnames(cordat[[i]]) <- labels
}


# put NA on diagonal of correlation matrix if variable is missing
for (i in 1:length(cordat)){
  for (j in 1:nrow(cordat[[i]])){	
    if (sum(is.na(cordat[[i]][j,]))==nvar-1) 
    {cordat[[i]][j,j] <- NA}
  }}

# show number of studies per correlation coefficient
pattern.na(cordat, show.na = FALSE)

# show total N per correlation coefficient
pattern.n(cordat, data$NUMBER.OF.BANKS)

# Run random effects model Stage 1
# For fixed effcets analysis change RE.type to "Zero"
stage1random <- tssem1(Cov=cordat, n=data$NUMBER.OF.BANKS, method="REM", RE.type="Diag")
summary(stage1random)

# Rerun model for convergence
stage1_rerun <- metaSEM::rerun(stage1random)

summary(stage1_rerun)

# model
model1 <- 
'# Regression coefficients
PERF ~ DUA+BNKRISK+BSIZE+LEV+AGE+LIQ+OWNCON+INST

# Covariances
DUA ~~ BNKRISK
DUA ~~ BSIZE
DUA ~~ LEV
DUA ~~ AGE
DUA ~~ LIQ
DUA ~~ OWNCON
DUA ~~ INST
BNKRISK ~~ BSIZE
BNKRISK ~~ LEV
BNKRISK ~~ AGE
BNKRISK ~~ LIQ
BNKRISK ~~ OWNCON
BNKRISK ~~ INST
BSIZE ~~ LEV
BSIZE ~~ AGE
BSIZE ~~ LIQ
BSIZE ~~ OWNCON
BSIZE ~~ INST
LEV ~~ AGE
LEV ~~ LIQ
LEV ~~ OWNCON
LEV ~~ INST
AGE ~~ LIQ
AGE ~~ OWNCON
AGE ~~ INST
LIQ ~~ OWNCON
LIQ ~~ INST
OWNCON ~~ INST
            
# Variances
DUA ~~ 1*DUA
BNKRISK ~~ 1*BNKRISK
BSIZE ~~ 1*BSIZE
LEV ~~ 1*LEV
AGE ~~ 1*AGE
LIQ ~~ 1*LIQ
OWNCON ~~ 1*OWNCON
INST ~~ 1*INST
PERF ~~ PERF'


RAM1 <- lavaan2RAM(model1, obs.variables=varnames)
RAM1

stage2 <- tssem2(stage1_rerun, Amatrix=RAM1$A, Smatrix=RAM1$S)

summary(stage2)