Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Different probability flows between simulations

Dear all!

To check parse behavior, I changed the value in overtaking.sumo.cfg to a invalid value:

...
<random_number>
       <random value='trsdue'/>
</random_number>
...

And included a debug WRITE_WARNING in OptionsCont::set() function (https://github.com/eclipse/sumo/blob/master/src/utils/options/OptionsCont.cpp#L249):
diff --git a/src/utils/options/OptionsCont.cpp b/src/utils/options/OptionsCont.cpp
index 8061dfd1e0..d2ebe7ff91 100644
--- a/src/utils/options/OptionsCont.cpp
+++ b/src/utils/options/OptionsCont.cpp
@@ -247,6 +247,7 @@ OptionsCont::getStringVector(const std::string& name) const {
 
 bool
 OptionsCont::set(const std::string& name, const std::string& value) {
+       WRITE_WARNING("set '" + name + "' = '" + value + "'");
     Option* o = getSecure(name);
     if (!o->isWriteable()) {
         reportDoubleSetting(name);
In sumo-gui directly, the tool tries to parse incorrect value and displays a error message (attached file):
image.png

but using TraCi it loads default value false to random (attached file):
image.png
Thanks in advance!

Em seg., 23 de nov. de 2020 às 09:49, Tiago Silva <tiago.silva1230@xxxxxxxxx> escreveu:
Hello!

I did some debugging yesterday and found that when I load the network via TraCi it uses the OptionsParser and when I load the network via sumo-gui it uses SAXParser. This is because myArgC is 3 via Traci and 0 via sumo-gui at this point in the code: https://github.com/eclipse/sumo/blob/master/src/utils/options/OptionsIO.cpp#L131

myArgV values using TraCi:
myArgV[0]: /home/tiago/src/sumo/bin/sumo-gui 
myArgV[1]: -c
myArgV[2]: overtaking.sumo.cfg

Does anyone know if this is the expected behavior? Do you know if the OptionsParser can parse random_number correctly?

Thanks in advance!

Em dom., 22 de nov. de 2020 às 11:39, Tiago Silva <tiago.silva1230@xxxxxxxxx> escreveu:
Hello everyone!

Forgiveness for the delay in returning. I performed some additional tests and found that the problem must be with the way I load the network. I noticed that if I load the network directly into the sumo-gui random works correctly. However if I use traci it is not working (which is my case). I use Omnet ++ with Veins. Therefore, the network is loaded via TraCi by Omnet ++.
I continue to investigate the reason for this. However, if anyone has any tips I would appreciate it very much.

Thanks in advance!

Em qua., 21 de out. de 2020 às 16:40, Michael Behrisch <oss@xxxxxxxxxxx> escreveu:
Hi,
I just tested the most recent build with --random and it works as expected.

Best regards,
Michael

Am 21.10.20 um 20:32 schrieb Tiago Silva:
> Dear Andras and Lokesh,
>
> Thanks for your reply! Unfortunately, even with seed and random my
> simulation is generating the same flow in different simulation runs. I
> tested this in v1_7_0 tag and in master branch with all current
> changes. Is this working for you? If so, it's probably a problem in my
> build...
>
> thanks in advance!
>
> Em qua., 21 de out. de 2020 às 13:56, Lokesh Das <lokeshcsedu@xxxxxxxxx
> <mailto:lokeshcsedu@xxxxxxxxx>> escreveu:
>
>     Hi Tiago,
>
>     From the heading, it seems that you want to change the vehicle
>     injection probability in every simulation run. If it is, you can do
>     it in two ways: i) manually- every time change probability value in
>     route file or ii) dynamically- change it using TraCI API. However,
>     if you want with the same probability every simulation should have
>     given you a different flow, just change the default seed value
>     (23423; see https://sumo.dlr.de/docs/Simulation/Randomness.html).
>     You can do it using <random value = "true" /> in *.sumocfg file. It
>     simply uses the system current time as a seed value. 
>
>     Thanks,
>     Lokesh
>
>     On Wed, Oct 21, 2020 at 8:15 AM Tiago Silva
>     <tiago.silva1230@xxxxxxxxx <mailto:tiago.silva1230@xxxxxxxxx>> wrote:
>
>         hello, 
>
>         I'm generating a flow of vehicles with probability='0.3' as seen
>         in the code below:
>
>         / <flow id='flow_2' color="0,0,1" begin='0' end='4500'
>         probability='0.3' type='vehicleType_1'>
>                   <route edges='5to1 1to3'/>
>         </flow>/
>
>         It works correctly and generates the random flows I need during
>         the simulation. However, I noticed that if I execute the
>         simulation again it will generate the same random flow as the
>         previous execution. Can you tell me if this is the expected
>         behavior of random flow? I need to generate different flows in
>         each simulation execution. Is there any way to do this?
>
>         Thanks in advance!
>
>         --
>         Att.
>         Tiago Alves Silva
>
>         _______________________________________________
>         sumo-user mailing list
>         sumo-user@xxxxxxxxxxx <mailto:sumo-user@xxxxxxxxxxx>
>         To unsubscribe from this list, visit
>         https://www.eclipse.org/mailman/listinfo/sumo-user
>
>     _______________________________________________
>     sumo-user mailing list
>     sumo-user@xxxxxxxxxxx <mailto:sumo-user@xxxxxxxxxxx>
>     To unsubscribe from this list, visit
>     https://www.eclipse.org/mailman/listinfo/sumo-user
>
>
>
> --
> Att.
> Tiago Alves Silva
>
>
> _______________________________________________
> sumo-user mailing list
> sumo-user@xxxxxxxxxxx
> To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
>


_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user


--
Atenciosamente,

Tiago Alves Silva



--
At.te,

Tiago Alves Silva

Analista de Desenvolvimento de Sistemas

Celular:+ 55 11 9 9819 7073

Skype: tiago.silva1230



--
At.te,

Tiago Alves Silva

Analista de Desenvolvimento de Sistemas

Celular:+ 55 11 9 9819 7073

Skype: tiago.silva1230

Attachment: Captura de tela de 2020-11-23 10-44-38.png
Description: PNG image

Attachment: Captura de tela de 2020-11-23 10-42-31.png
Description: PNG image


Back to the top