Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Introducing generics #44

Open
wants to merge 9 commits into
base: master
Choose a base branch
from
Open

Introducing generics #44

wants to merge 9 commits into from

Conversation

Ploppz
Copy link

@Ploppz Ploppz commented Feb 2, 2020

Discussion: #36

TLmaK0 and others added 9 commits February 1, 2020 16:28
@TLmaK0 @Ploppz
Able to see function approximation
d63104f
@TLmaK0 @Ploppz
Added bias
2155ec7
@TLmaK0 @Ploppz
Fixed ctrnn
5ad940e
@TLmaK0 @Ploppz
Fixing problems with activation operations
210a619
@TLmaK0 @Ploppz
Added exception when fitness less than 0
9f9fe3c
@TLmaK0 @Ploppz
Remove inject_gene calls
287d8bc
@Ploppz
Generic rewrite
9cf2576 
- Genome is a trait that implements mutation, mating, distance measure etc.
        (opens up for more experimentation and extensibility)
- The default Genome implementor is NeuralNetwork which implements said functionality
        pretty much just as before. So NeuralNetwork is the new Genome.
        Some slight changes in how bias works had to be done to be able to mirror the way
        they do it in some paper that introduced or explained CTRNN iirc.
- Thus now we pass around Organisms not Genomes

- And..:
    - Update rand
    - Removed lots of unnecessary f64
    - Remove CtrnnNeuralNetwork (can just use Ctrnn for setting up the
        network)
    - Fix an error in species
@Ploppz
Cargo fmt and remove the 'can solve xor' test for now
9be47b5
@Ploppz
fix travis?
3339726
@TLmaK0
Copy link
Owner

TLmaK0 commented Feb 9, 2020 •
edited
Loading

We should remove any commit before "Generic rewrite" not included in master.

@Ploppz
Copy link
Author

Ploppz commented Feb 9, 2020

I can try to revert those commits.
I do think we eventually want bias, seeing as it is introduced in the paper, and other implementations like python-neat also include it.
In my final version I reached a good performance with the help of bias.
In the next PR I reworked the bias system actually: I introduce genes for neurons, which hold the bias. It seemed more logical than having an is_bias flag for connections.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@Ploppz @TLmaK0