genetic optimization looks like exhaustice  [SOLVED]

Questions about MultiCharts and user contributed studies.
milda8005
Posts: 13
Joined: 09 Aug 2018
Been thanked: 1 time

genetic optimization looks like exhaustice

Postby milda8005 » 13 Sep 2018

hi when i try to optimize the strategy i would like to use a genetic optimization becouse of no of variations. but if i set genetic, the number of combinations looks the same as exhaustive, i have read that this "bug" is solved in version 12 but it seems that its not. see screenshot
then it takes 5 days in stead of hours
Attachments
genetic_02.JPG
(25.42 KiB) Downloaded 498 times
genetic_01.JPG
(45.68 KiB) Downloaded 498 times

User avatar
Anna MultiCharts
Posts: 461
Joined: 14 Jul 2017
Has thanked: 31 times
Been thanked: 113 times

Re: genetic optimization looks like exhaustice

Postby Anna MultiCharts » 14 Sep 2018

Hello, Milda8005!

Could you please be more specific?
Would you like to know why the number of simulations with modified settings for parameters “Set Population Size” and “Maximum Number of generations” doesn’t change compared to the number of simulations with default parameters settings? Or do you want find out why the number of combinations is this much different between exhaustive and genetic optimizations: 10179487334016 vs 1315242? Or maybe you have something else in mind?

milda8005
Posts: 13
Joined: 09 Aug 2018
Been thanked: 1 time

Re: genetic optimization looks like exhaustice

Postby milda8005 » 15 Sep 2018

those different numbers are i think not a deal, but if I do the same setup 10 population and 1000 generations - it is in overall 10000 test to be done and not over 1 million, because with the same settings in TS it is so and in TS it is done within hours but if i will put over 1 million in TS it takes although the 5 days to done.
main issue what i mean that i am not sure if i set 10 population and even 100 generation as it is in screenshot above does not make 1milion in genetic optimization.
hope it is more clear now
thanks

milda8005
Posts: 13
Joined: 09 Aug 2018
Been thanked: 1 time

Re: genetic optimization looks like exhaustice  [SOLVED]

Postby milda8005 » 19 Sep 2018

Solved b last upsate

User avatar
Anna MultiCharts
Posts: 461
Joined: 14 Jul 2017
Has thanked: 31 times
Been thanked: 113 times

Re: genetic optimization looks like exhaustice

Postby Anna MultiCharts » 21 Sep 2018

milda8005,

Indeed in MultiCharts 12 Release 1 there was a bug when input values didn’t influence the number of simulations in basic genetic optimization. This bug was fixed in MultiCharts 12 Release 2.
But there still remains a bug in basic genetic optimization with the number of simulations being calculated incorrectly when Set Population Size < 100. This is expected to be fixed in the nearest releases.


Return to “MultiCharts”