Re: [sumo-user] Distribution of vehicle types with vehTypeDistribution.p

[Jakob Erdmann <namdre.sumo@xxxxxxxxx>]

This is not supported by createVehTypeDistribution.py

Also, you would get problems with correlated parameters (supposedly trucks are longer on average than passenger cars) and this could not be handled by the tool.

I recommend creating 3 distributions with different base probabilties by like this:

...

vClass; passenger

probability; 90

--------

...

vClass; truck

probability; 8

---------

...

and then merging them all into a single distribution. (could be done one a unix shell using  'head', 'tail', and 'cat')

regards,

Jakob

Am Di., 5. März 2019 um 11:53 Uhr schrieb Fredrik Nilsson <frnilss@xxxxxxxxxxxxxxxxxxx>:

Hi,

I am using vehTypeDistribution.py to generate a population of vehicles in my simulation and I was wondering how I could make it generate different vehicles types (with a certan distribution, e.g. 90% passenger cars, 8% trucks and 2% buses)? I currently use a config file like this (don't mind the numbers they are in the middle of running an optimization):

accel; normal(1.4999999999936855, 0.4999999999976304); [0.5, 1.5]

decel; normal(5.999999999999464, 0.7999999999995001)

sigma; normal(0.3000000000017896, 0.1999999999983296); [0.1, 1]

tau; gamma(18.198814787804807, 4.00000000000155); [0.5, 6]

length; normal(4.9, 0.2); [3.5,5.5]

minGap; normal(3.4514375986324963, 0.1000000000006245)

lcStrategic; normal(7.999999999999491, 0.8123908338568304)

lcCooperative; normal(0.10000000000437191, 1.0); [0.1, 1]

vClass; passenger           <------------------ Change something here I assume?

carFollowModel; Krauss

BR,

Fredrik

_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/sumo-user