search

CityScope / CS_Simulation_GAMA

Agent Based Simulation platform for CityScope
GNU General Public License v3.0
31 stars 21 forks source link

Autonomy Vs Conventional: #95

Closed agrignard closed 4 years ago

agrignard commented 4 years ago

I think that it is confusing to have a tag that jumps from the scenario of Autonomy to Conventional without re-starting the simulation. Either we restart the simulation with that tag, or we rune both simulation at the same time and we visualize one or the other with the tag.

agrignard commented 4 years ago

I just add a new experiment called MultiScenarios with the 2 scenarios in //

Screenshot 2020-04-30 at 11 50 00
LAAP commented 4 years ago

Hi @agrignard ,

Great! There are some issues:

1) Now we are missing the Graph with infected people etc.

2) The speed of the simulation, only affects to the autonomous scenario (the graph in the right):

Captura de pantalla 2020-04-30 a las 13 14 45

3) When you change the tag from Conventional or autonomy and vice versa, only affects the to the scenario in the right:

Captura de pantalla 2020-04-30 a las 13 11 31

I suggest either keeping just the autonomous scenario with the tag, or the 2 scenarios and eliminate the tag

agrignard commented 4 years ago

The problem here that i d'int think about before to rpopse youa multi experiment is related to the way GAMA is made: any parameters only have an impact on the last (so the current) describe the experiment. So I don't think it's the right solution to have two experiments here .

The question is do we need two experiment in // or can't we produce 2 videos? In that case it's 2 separated experiment.

LAAP commented 4 years ago

Hi @agrignard ,

I think that there are 2 options:

1) 2 experiments in the same model, that shifts from one to the other (visually but also computationally) son they run only when the "autonomy tag" is on or off, so they restart everything when you play with the "tag". In this option, maybe the "infection Graph" can stay as a benchmark in a very light color in the back, so you can see the difference between 1 model and the 2nd

2) Two separate models. I think that this is the easier one

Both options are good to me

agrignard commented 4 years ago

Sorry, I don't understand the issue here and I think we do'nt have the same definition of model and experiments. The first option that you mention here is what was already implemented here. When you run the model you have a button (Is this what you call "tag"?) to switch between conventional and autonomy

The option 2 doesn't make sense for me to have 2 separate models are the 2 models are very similar

agrignard commented 4 years ago

Going back to the beginning with only one scenario...