pq=p+q
x1=x-mean(x)
n=dim(x1)
coef=matrix(rep(0,pq*pq),pq,pq)
for (i in 1:pq){
 md=arma(x1,order=c(i,0), include.intercept=F)
 for (j in 1:i){
 coef[i,j]=md$coef[j]
 }
 }

eacf=matrix(rep(0,p*q),p,q)
m1=acf(x1,lag.max=q,plot=F)
for (j in 1:q){
eacf[1,j]=m1$acf[j+1]
}


for (j in 1:q) {
tmp=coef
 for (i in 1:(pq-j)){
  for (ii in 1:i){
   if (ii == 1) coef[i,ii]= tmp[i+1,ii]+tmp[i+1,i+1]/tmp[i,i]
   if (ii >=2) coef[i,ii]=tmp[i+1,ii]-tmp[i+1,i+1]*tmp[i,ii-1]/tmp[i,i]
}
}

for (i in 2:p){
y=matrix(rep(0,n[1]),n[1],1)
for (it in i:n[1]) {
  w=x1[it,1]
  for (ii in 1:(i-1)){
   w=w-coef[(i-1),ii]*x1[it-ii,1]
  }
  y[it,1]=w
}
 m1=acf(y,lag.max=q,plot=F)
 eacf[i,j]=m1$acf[j+1]
}
}

seacf=matrix(rep(0,p*q),p,q)
sd = 2/sqrt(n[1])
for (i in 1:p){
 for (j in 1:q){
  if (abs(eacf[i,j]) >= sd) seacf[i,j]=2
}
}

print('EACF table')
print(eacf)
print(' ')
print('Simplified EACF: 2 denotes significance')

print(seacf)