Hi,

I have tried to use initial cluster centers from hierarchical clustering (an external file) when doing k-means analysis. However, I always reiceive an error message. Although I do use the correct format "cluster_" I still get an error message when I read in the external file saying "The file referenced on the FILE subcommand does not contain initial cluster centers for at least one of the variables being analyzed." I am not sure how to proceed as all of my variables do contain cluster centers. Do you know what I can do to make it work?

I can't guess what is wrong with your centers file, but I suggest running a k-means procedure and saving the centers file from it. Compare that with what you are using as input to see if you can spot the difference in the variable dictionaries.

Hello, I meet the same case. I want read in the initial cluster center， but the file is the wrong foamat. Could you tell me the right file format?

Hi Jon, thank you for your response! I have actually tried that and discovered no difference. I also tried using other data, another laptop, even the spss subscription just to see if I can get it to work but nothing helped. I wonder if it is an spss bug and just does not work. Or does it work when you try it? Then it is just me.

I have used this successfully many times. If you want to send me the data or a sample of it, the centers file, and the syntax, I'll take a look. (jkpeckSupport Member)

Jon, thank you soo much! I have just sent you an email and hope you received it and all the files. I am curious to know what you can find out.

Hello, I meet the same case. If you had solved it, can you give me some advise?