Hi all,
Does anyone knows how to handle ordered preferences applying the R
package mlogit (multinomial logit model)? My data set provides for each
customer preferences (given as percentages) for 6 different brands. I
would like to use for model calibration not just that brand with
maximum stated preference.
I know that ordered preferences can be used in the R package MNP
(multinomial Probit model) but I can't use that package due to numerical
problems. Mlogit has an option "weights". Can that option be used to
solve my problem?
Best regards,
Holger