Hey,
In frame of a cluster analisis I'd like to detect similarities between some objects. The varibles included show diffrent scales (categorial and binary). I've transformed the categorial ones in dummy-variables. To avoid a distortion of the results caused by this transformation, I'd like to weight the ex-categorial - now dummy-variables with factor (1/number of categories).
1) At which point of the cluster analysis I've to cunduct the weighting? (Data preperation? clustering?)
2) Which command structure/syntax do I have to use?
Thanks for your hints and help!
In frame of a cluster analisis I'd like to detect similarities between some objects. The varibles included show diffrent scales (categorial and binary). I've transformed the categorial ones in dummy-variables. To avoid a distortion of the results caused by this transformation, I'd like to weight the ex-categorial - now dummy-variables with factor (1/number of categories).
1) At which point of the cluster analysis I've to cunduct the weighting? (Data preperation? clustering?)
2) Which command structure/syntax do I have to use?
Thanks for your hints and help!