topic clustering continuous variables in Statistical Procedures https://communities.sas.com/t5/Statistical-Procedures/clustering-continuous-variables/m-p/55359#M2541 <HTML><HEAD></HEAD><BODY><P>Hi, I know you can't cluster by continuous variables, but looking for ways around it.. For example I have price1, price2 and price3&nbsp; on a monthly intervals for the last 15 years of about 120 stores.&nbsp; I want to cluster stores that have some sort of "something" in common between the 3 prices.&nbsp; But do I look at average prices, avergae growth/loss, price growth in the last 5 years, slope, etc. or a combination .&nbsp; Looking to see what makes most sense and which clustering procedure works best for this.&nbsp; Thanks</P></BODY></HTML> Mon, 28 Nov 2011 16:34:36 GMT podarum 2011-11-28T16:34:36Z clustering continuous variables https://communities.sas.com/t5/Statistical-Procedures/clustering-continuous-variables/m-p/55359#M2541 <HTML><HEAD></HEAD><BODY><P>Hi, I know you can't cluster by continuous variables, but looking for ways around it.. For example I have price1, price2 and price3&nbsp; on a monthly intervals for the last 15 years of about 120 stores.&nbsp; I want to cluster stores that have some sort of "something" in common between the 3 prices.&nbsp; But do I look at average prices, avergae growth/loss, price growth in the last 5 years, slope, etc. or a combination .&nbsp; Looking to see what makes most sense and which clustering procedure works best for this.&nbsp; Thanks</P></BODY></HTML> Mon, 28 Nov 2011 16:34:36 GMT https://communities.sas.com/t5/Statistical-Procedures/clustering-continuous-variables/m-p/55359#M2541 podarum 2011-11-28T16:34:36Z