DataFrames and Models

[Matt Weiden]

Hello!

Though the use of DataFrames with factors in fitting linear models is mentioned in the doc

https://github.com/HarlanH/DataFrames.jl/blob/master/doc/sections/03_design_details.md#factors

I get an error when trying to fit a linear model on a data frame with (any) factors in it. It seems that there should be some functionality in the linear model code to expand the factors into indicator features automatically or a method to expand the DataFrame with factors in it into a new DataFrame (or fake it) such that it can be fit.

Could you provide an example of this use case?

Thanks,

Matt

[John Myles White]

Doug recently changed this functionality and I haven't had a chance to catch up with his work. Hopefully he'll chime in.

Could you switch this conversation to the julia-stats mailing list?

-- John

[Douglas Bates]

On Wednesday, July 10, 2013 5:28:31 PM UTC-5, John Myles White wrote:

Doug recently changed this functionality and I haven't had a chance to catch up with his work. Hopefully he'll chime in.

Could you switch this conversation to the julia-stats mailing list?

Please do.  Also, can you give an example of what happens?  Are you using the lm function from the GLM package?