I'm attempting to create my own function that uses the box cox transformation and returns all values of lambda in which the new data is normally distributed. Not sure where I have gone wrong. This is my code:

boxtran<-function(x){

lambda=seq(-5,5,1/10)

for(i in lambda) {

if (i!=0){

y<-(x^i-1)/i

}

if (i==0){

y<-log(x)

}

p<-ad.test(y)$p.value

}

q[p<=0.05]

for (seq(min(q),max(q))){

return(i)

}

}

}