Cluster Analysis of Panel Data: Possible?

[Zia Sani]

Hi,

I'm working with a panel data set of below features:
* gvkey: 989
* year: 2006-2016 (without gaps)
* Unbalanced

I want to perform a cluster anaylsis, using K-means partition-clusteting method, based on 4 variables to divide firms into 4 groups. My question is that, does Stata command -cluster- apply to panel data? In other words, can I perform cluster analysis of panel data in Stata?

As I understood from cluster analysis literature and Stata manuals that cluster analysis is about defining groups in data as it assigns "observations" to closest cluster applying a criteria ex. Distance measure, where analysed data is of cross-section form.

Thank you for your assistance.

[River Huang]

Hi, Zia: I wonder if the following paper helpful. Please have a look. https://sites.google.com/site/kerryd...me/stata-files.

[Zia Sani]

Hi, Zia: I wonder if the following paper helpful. Please have a look. https://sites.google.com/site/kerryd...me/stata-files.

Hi River,

Thanks a lot for sharing.