🔒 reinforcement learning (NEAT) help

[hardy grotto]

@visual pike

Python help channel opened

Remember to:

• Ask your Python question, not if you can ask or if there's an expert who can help.
• Show a code sample as text (rather than a screenshot) and the error message, if you've got one.
• Explain what you expect to happen and what actually happens.

:warning: Do not pip install anything that isn't related to your question, especially if asked to over DMs.

Closes after a period of inactivity, or when you send !close.

[visual pike]

bro wtf

[hollow hull]

You can't upload files here

[visual pike]

how can i send my code then

[hollow hull]

If you'd like to share your code, you'll have to use a pastelink

!paste

[hardy grotto]

Pasting large amounts of code

If your code is too long to fit in a codeblock in Discord, you can paste your code here:
https://paste.pythondiscord.com/

After pasting your code, save it by clicking the Paste! button in the bottom left, or by pressing CTRL + S. After doing that, you will be navigated to the new paste's page. Copy the URL and post it here so others can see it.

[visual pike]

i tried

i sent the link for it

https://paste.pythondiscord.com/VSPA

there now its not deleted

[visual pike]

guys i need ur opinion on my code wether i did any mistakes or not, cuz the code doesnt work (i gave it like 20mins training time)

[sour fog]

what results are you seeing?

like ofc it doesn't work but how is it failing

score never goes up?

[visual pike]

let me record it for u

[visual pike]

i made the score based on how close the net outputs to the pid controller

u see how it first falls like that then it just keeps falling as so

what is the diffirence between fitness and adj fitness?

[sour fog]

I don't see adjusted fitness or that string 'adj fit' mentioned anywhere in code you sent so I can't say

I would try printing out the outputs of your model to see if it's an issue with your simulation code

if the model is giving different outputs but it's still doing the exact same thing on screen there's an issue with the simulation

otherwise if the models are not changing at all and are giving the exact same output I would print out your scores and take a look at the structures of the models after you run NEAT each time to see if they are actually changing

it's hard to be any more specific since I don't have the assets to run the sim on my own machine without spending the time to dig through the code removing all of the rendering and file loading

[visual pike]

first: sry i was afk and in the mean time i left the code training, second: the printed data on the screen is from the library called "neat" in python

the simulation is fine i think, since i tested it with player mode and pid mode, and they both work well, all i did to make the ai control the drone is i made it change the drone thrust and stuff

also it seems to be develeping some kind of path it always does every time it respawns

Can u suggest me how can i train it, like the starting conditions and a way to make learn to fly to targets and if i should have some kind of randomness in the starting conditions and so on. thx

[hardy grotto]

Python help channel closed

This help channel has been closed and it's no longer possible to send messages here. If your question wasn't answered, feel free to create a new post in #1035199133436354600. To maximize your chances of getting a response, check out this guide on asking good questions.