Destination choice model

137 views
Skip to first unread message

Daros Kpikpidi

unread,
Nov 17, 2021, 12:20:46 AM11/17/21
to Biogeme
Hello Prof
I would like to have some of your advice on using biogeme for a model. Indeed,
I want to use Biogeme to extimate a model of choice of destination, that is to say a specification of the form Vij = Beta_duree * Duree_ij + Beta_distance * Disatance_ij + log {population_j + Beta_emploi * Emploi_j}.
with i the origin and j the destination

I have 40 TAZ (Traffic analysis zone) and in the literature it says to make a random choice of a small number of alternatives for each observation. Let's oppose that for each observation I want to consider 20 alternatives, how to do it with Biogeme?

In addition, with biogeme, it will be necessary to have one column per variable (in the model) according to each alternative. I do not have in my base. since if I take a trip from an origin 1 to a destination 6, we can only have the duration, the distance for this only alternative. I don't have the duration and the distance of 1 to the other 39 alternatives. How to proceed please?

Bierlaire Michel

unread,
Nov 17, 2021, 12:41:58 AM11/17/21
to acalm...@gmail.com, Bierlaire Michel, Biogeme

On 17 Nov 2021, at 00:15, Daros Kpikpidi <acalm...@gmail.com> wrote:

Hello Prof
I would like to have some of your advice on using biogeme for a model. Indeed,
I want to use Biogeme to extimate a model of choice of destination, that is to say a specification of the form Vij = Beta_duree * Duree_ij + Beta_distance * Disatance_ij + log {population_j + Beta_emploi * Emploi_j}.
with i the origin and j the destination

I have 40 TAZ (Traffic analysis zone) and in the literature it says to make a random choice of a small number of alternatives for each observation. Let's oppose that for each observation I want to consider 20 alternatives, how to do it with Biogeme?

It is just a data preparation issue. You need to prepare your data file using Pandas, Excel, or whatever database software that you like.


In addition, with biogeme, it will be necessary to have one column per variable (in the model) according to each alternative. I do not have in my base. since if I take a trip from an origin 1 to a destination 6, we can only have the duration, the distance for this only alternative. I don't have the duration and the distance of 1 to the other 39 alternatives. How to proceed please?

Well, like for any revealed preference data, you’ll have to impute the variables for the unchosen alternatives. 


--
You received this message because you are subscribed to the Google Groups "Biogeme" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeme/1812d841-0088-4e3f-a3a4-af5a831e7d71n%40googlegroups.com.

Daros Kpikpidi

unread,
Nov 17, 2021, 12:47:03 AM11/17/21
to Biogeme
thank you very much Prof
Reply all
Reply to author
Forward
0 new messages