In storage.mode(x) <- "double" : NAs introduced by coercion in kmeans

Question

I am trying to split my data into 5 clusters. But I am getting the following error

> colSums(sapply(train1,is.na))
     train_id              name item_condition_id     category_name 
            0                 0                 0                 0 
   brand_name             price          shipping  item_description 
            0                 0                 0                 0 
> train1matrix=as.matrix(train1)
> train1vector=as.vector(train1matrix)
> k=5
> set.seed(88)
> KMC=kmeans(train1vector,centers=k,iter.max=1000)
Error in do_one(nmeth) : NA/NaN/Inf in foreign function call (arg 1)
In addition: Warning message:
In storage.mode(x) <- "double" : NAs introduced by coercion

Can someone please help me? Thank you in advance

Solution

kmeans can only be used on numerical columns, because it needs to compute the mean.

Don't use it on "ID" columns, text columns etc. where it does not make sense to compute the mean. It appears that you are trying to run kmeans on such 'bad' columns.

Comments

[MKR]

Please use fput to share data in train1.

[Srihari Mohan]

Thank you... But I am yet to try..do you mean to tell that to copy data from old data frame to new dataframe, I shud use fput?

[Marco Sandri]

You should share the output of dput(train1) or, at least dput(train1[1:20,])