Hello Statalisters,
I am running a cluster analysis using Ward method after which I am trying to create grouping variables for: say, four clusters. Due to ties I have to specify the ties(more) option: cluster gen group = groups(4), ties(more). When I run tab group I see that six groups have been created but one or two of them contain no observations. I don´t fully understand why this has happened. So far, I have chosen to ignore the empty clusters and just work with the remaining clusters but am not confident if this is legitimate.
The dataset I am using contains sequence data and i am running the cluster analysis on a full distance matrix obtained by the optimal matching algorithm.
Best regards,
Bernd
I am running a cluster analysis using Ward method after which I am trying to create grouping variables for: say, four clusters. Due to ties I have to specify the ties(more) option: cluster gen group = groups(4), ties(more). When I run tab group I see that six groups have been created but one or two of them contain no observations. I don´t fully understand why this has happened. So far, I have chosen to ignore the empty clusters and just work with the remaining clusters but am not confident if this is legitimate.
The dataset I am using contains sequence data and i am running the cluster analysis on a full distance matrix obtained by the optimal matching algorithm.
Best regards,
Bernd
