R:data simulation - r

Question

I am m trying to run the following code:
data_greene<-read.delim(file.choose(),header=T)
result_b2_HC0<-matrix(1:2000,ncol=4)
for (i in 1:500){
X1<-data_greene[[3]]*10^-4
X2<-X1^2
e<-rnorm(50,0,1)
sigma2<-exp(5.30+5.30*X1)
lambda<-max(sigma2)/min(sigma2)
Y<-1+1*X1+0*X2+sqrt(sigma2)*e
lms<-lmsreg(Y~X1+X2)
yhat<-lms$fitted
resid<-lms$residual
s<-abs(resid)
lms2<-lmsreg(s~yhat)
shat<-lms2$fitted
w1<-1/shat^2
scale<-lms$scale[1]
stdres<-resid/scale
e=abs(stdres)
w2<-NULL
for (i in 1:50){
if(e[i]<=1.345) w2[i]<-1 else w2[i]<-1.345/e[i]
}
w<-w1*w2
WLS<-lm(Y~X1+X2,weights=w)
res1<-WLS$residual
HCCMEHC0<-function(Y,X1,X2){
X<-cbind(1,X1,X2)
W<-diag(w)
inv<-solve(t(X)%*%W%*%X)
psi0<-diag(res1^2)
HC0<-inv%*%t(X)%*%W%*%psi0%*%W%*%X%*%inv
return(HC0)
}
result_b2_HC0[i,1]<-WLS$coef[3]
result_b2_HC0[i,2]<-sqrt(HCCMEHC0(Y,X1,X2)[3,3])
result_b2_HC0[i,3]<-result_b2_HC0[i,1]/result_b2_HC0[i,2]
result_b2_HC0[i,4]<-2*pt(-abs(result_b2_HC0[i,3]),df=47)
}
result_b2_HC0
I would expect the matrix to be complete, but the result only appears at row 50 in the matrix. What am I doing wrong?

Answer 1

You are using the same variable i in two nested for loops. Change the second for loop to use the variable j instead.
To avoid this error, make sure you always use indentation. Also, learn how to use vector mathematics. Your second loop can be rewritten from
e=abs(stdres)
w2<-NULL
for (i in 1:50){
if(e[i]<=1.345) w2[i]<-1 else w2[i]<-1.345/e[i]
}
to
e=abs(stdres)
w2<-ifelse( e <= 1.345, 1, 1.345/e )
This is cleaner, easier to read, and faster.