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.

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

More tests! #53

Open
2 tasks
shiffman opened this issue Feb 10, 2018 · 9 comments
Open
2 tasks

More tests! #53

shiffman opened this issue Feb 10, 2018 · 9 comments

Comments

@shiffman
Copy link
Member

Neural Network class needs tests?

  • test feed forward?
  • test backprop?
@AR-234
Copy link

AR-234 commented Feb 11, 2018

One problem with that is that node needs to import the matrix lib into nn and i haven't found a way to bypass the browser error with the require statement..
otherwise i would be happy to do some tests.

@shiffman
Copy link
Member Author

Well I guess this should eventually move to a more proper build system, but for now would a solution like I'm doing here work, but with require instead?

if (typeof module !== 'undefined') {
  module.exports = Matrix;
}

@MTRNord
Copy link
Contributor

MTRNord commented Feb 11, 2018

There is also still the solution by me in #13 to rely on ES6 which plays fine with browsers. It would remove the if completly (but has as stated in that PR other side effects)

@shiffman
Copy link
Member Author

Thank you for this reminder @MTRNord, apologies I haven't had the time yet to learn more about ES6 modules and review in more detail.

@MTRNord
Copy link
Contributor

MTRNord commented Feb 11, 2018

@shiffman no problem :) take your time to learn it 👍 It isn't urgent to have it but it is a "Nice to have" from my perspective :)

@AR-234
Copy link

AR-234 commented Feb 11, 2018

ah ok, awesome, and the typeof approach doesn't work, because Matrix is defined
but i guess i will try the ES6 thing

@MTRNord
Copy link
Contributor

MTRNord commented Feb 11, 2018

@AR-234 keep in mind what I wrote in my PR the ES6 way doesn't work "out of the box" until the PR is merged. (thats the side effect)

@AR-234
Copy link

AR-234 commented Feb 11, 2018

but maybe there should be a building process: if the lib keeps expanding and stuff like a class for training methods get added, or even more advanced stuff, it could get really bad with the readability of the file

@Adil-Iqbal
Copy link

I remember something being mentioned about the Neural network library being expanded to include a flexible number of hidden layers. If that's true, it would help to know what that looks like before tests are written. There are a few ways it could be implemented.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants