问题
I am running a logit regression in R with a large number of input variables.
newlogit <- glm(install. ~ SIZES + GROSSCONSUMPTION.... +
NETTCONSUMPTION..... + NETTGENERATION....... +
GROSSGENERATION.... + Variable. + Fixed +
Cost.of.gross.cons + Cost.of.net.cons + Cons.savings +
generation.gains + Total.savings + Cost.of.system +
Payback + Self.consumption + Total.consumption.as.solar +
Owner.occupied + postcode + Suburb + Market.penetration +
X..green.vote, data = newdata, family = "binomial")
I am getting this error:
Error in contrasts<-(tmp, value = contr.funs[1 + isOF[nn]]) : contrasts can be applied only to factors with 2 or more levels
Would you please be able to advise why this error is occurring and if there is a way to run this regression?
回答1:
It objects to one of your variables (as gvrocha said); you might have a factor with only one level, or a string.
A tip to quickly track down the offending variable(s) is to do interval bisection and increase/decrease the col indices till you trigger the error.
Best to use the numerical (column-index) interface to glm (glm(data[,'install.'] ~ data[,2:40]), see [1])
rather than the formula interface glm(install. ~ var1 + var2 + ...:
[1] Dynamic formula creation in R?
回答2:
I've seen this problem happen when one of the variables on the right hand side of the model is a factor with a single level.
If that is the case, all you need to do is to remove that factor from the model.
来源：https://stackoverflow.com/questions/29980426/regression-gives-error-on-one-of-the-input-variables-contrasts-can-be-applied-o