CODEWITHSUNDEEP
rregressionglmmodel-comparison
user3391456
user3391456

Reputation: 21

Error in eval(expr, envir, enclos): no loop for break/next, jumping to top level

I am trying to run a stepwise analysis on a negative binomial model with many factors:

step(glm.nb(totphy~lat+long+elev+cult+poll+blkage+canden+irr+dust+dustcon+mid+herb+miteprob+ermprob+tsmprob+mcdprob+brprob+armprob+mitapp+acra+agri+carz+envi+fujim+next+vend+zeal+cover+orchtype+pestint+mating+ties+rimon+igr+lime+naa+sevin+land+area+ermpres+tsmpres+mcdpres+armpres+zmalipres+ltemp+lmaxtemp+lmintemp+ldewp+lrh+lwindsp+lwinddir+lsolar+lprecip+lleaf+lsoil,data=mites), direction="forward", criterion=aic)

This code works if I only include the variables up to "fujim", adding more variables causes this error:

Error in eval(expr, envir, enclos) : 
  no loop for break/next, jumping to top level

I am fairly unfamiliar with R, so what does this mean?

Upvotes: 2

Views: 1960

Answers (1)

epsilone
epsilone

Reputation: 765

As Karl Forner points, the problem is probably on a variable called "next", so just rename it. Check the following code:

# Data
data=data.frame("y"=rnorm(10),"x"=rnorm(10),"next"=1:10)

# Error
lm("y~x+next",data=data)

# Good
colnames(data)[3]="mod.next"
lm("y~x+mod.next",data=data)

Upvotes: 3

Related Questions