Cheers, ad

]]>I will try to explain it as simply as I can (because it sounds complicated in my head)

Problem: I don’t have variables for which I want to find correlations hanging around in my global environment. I have data-frames. The particular one I am using now for the regression is called regdata. I am modeling my lm regression like this. The size of the dataframe is 160 x 9, 160 rows and 9 columns.

result 2″ to an “invalid object”. I have tried all of the following and nothing works

summary(result, cluster = c (regdata$x3))

summary(result, cluster = c (“regdata$x3”))

summary(result, cluster = c (x3))

summary(result, cluster = c (“x3”))

summary(result, cluster = c (160, regdata$x3))

In this instance, x1, x2, x3 are all categorical variables with

x1 ranging from 1 to 5

x2 has 3 values 0,1,2

x3 has 4 values ranging from 1 to 4

Can I not cluster if the number of clusters in more than 2?

]]>