pyMVPA: how to regress out covariance (poly_detrend)

Dear pyMVPA experts,
I want to regress out participants’ willingness to pay ($0, 1,3,5) from the brain response which is a different attribute than chunks (run number 1-6 (2runs*3 days)). Based on the information here http://www.pymvpa.org/generated/mvpa2.mappers.detrend.html#module-mvpa2.mappers.detrend, I want to use this line: poly_detrend(dataset, polyord=1, chunks_attr=‘wtp’). Should I add the space parameter since my chunks_attr= wtp is not ordered from low to high?

Thanks for your help.