Error in Oversampling in R

I am working on SMOTE using DMwR package. i tried oversampling of data but i got error saying The response variable must have 2 levels.
what does it mean ? How can i solve this error ?
i tried following code:

> over <- ovun.sample(tumor_stage~., data = train, method = "over", N = 760)$data

got error

Error in (function (formula, data, method, subset, na.action, N, p = 0.5,  : 
 The response variable must have 2 levels.````

how many levels are there in train$tumor_stage ?

in train$tumor_stage has 10 levels

as an experiment, if you recoded it down to two levels, would the function run without error ? it does seem to me to be complaining about your response variable having not exactly two levels.

with 2 levels it works..

If you need to model 10 levels, you will need an appropriate tool for the job. Perhaps others here have suggestions

1 Like

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.