I am trying to run a mlogit
in R using a travel diary of ridehailing trips. The dependent variable is solo vs. pool (e.g., UberX vs. UberPOOL) (See screenshot of the data after applying mlogit.data
. Each person (id) did different numbers of trips. Age and Income are individual-specific variables, "Taxi_type_payment" (how the person paid the ride) is trip-specific, and "DUR" is the duration of the trip.
f <- mFormula(choice ~ DUR | HH_income + Age +Taxi_type_payment | 0) f1 <- mFormula(choice ~ 0|Taxi_type_payment + HH_income + Age + DUR|0) ml <- mlogit(f1, data, reflevel="1")
mlogit
runs with mformula f1
and does not with formula f
. f
gives the following error:
Error in solve.default(H, g[!fixed]) : system is computationally singular: reciprocal condition number = 7.50953e-17
I would appreciate any help to understand why.
Thank you.
https://stackoverflow.com/questions/67410592/error-in-solve-defaulth-gfixed-system-is-computationally-singular-recip May 06, 2021 at 09:01AM
没有评论:
发表评论