ERROR: MethodError: no method matching polynomial_ring(::ZZRing, ::Tuple{Symbol, Symbol}) about codingtheory HOT 12 CLOSED

Yeah, this is due to an update in a dep. I fixed it the other day on the dev branch, which also contains some fixes for other bugs introduced by the update. Please use dev until merged into master.

By examples you are referring to the documentation? That takes significant time to write and hasn't been done yet. On the list...

from codingtheory.

I get the error on dev too:

julia> regular_LDPC_code(20, 20, 20)
ERROR: MethodError: no method matching polynomial_ring(::Nemo.ZZRing, ::Tuple{Symbol, Symbol})

I think I'm using dev, tried both Pkg.add(url="here#dev") and rev="dev" arguments.

At least something changed, I'm not seeing GF defined anymore.

from codingtheory.

Huh... that polynomial ring should be univariate. It sounds like you are on branch up_deps. GF is still in dev and it should be PolynomialRing instead of the new polynomial_ring in up_deps.

julia> regular_LDPC_code(20, 20, 20)
[20, 19, 2]_2 regular (20, 20)-LDPC code with density 1.0.

Variable degree polynomial:
        1//20*x^20
Check degree polynomial:
        1//20*x^20
Parity-check matrix: 20 × 20
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
        1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

This function is out of a textbook, but I have not written any unit tests for it, so let me know if there's anything weird about it. You may also submit tests...

Either way, when I finish with getting the tests to recognize the new extension I made, I will pull dev into master to fix this.

from codingtheory.

Nope, went as far as cloning to a local repo, merging dev into master, deleting every other branch and adding it via Pkg.add(path), still getting the same error. Is Pkg pulling some old dependencies?

from codingtheory.

I've merged the fixes from dev into master. I'd say delete, try master again, and if it fails, please post the full stack trace.

from codingtheory.

On a fresh install with empty .julia directory. Full log attached: logfile.txt

from codingtheory.

From your log it seems you're not using the Manifest.toml. Make sure you haven't updated that file and run instantiate in the package REPL from within the project environment.

from codingtheory.

I was trying to find how to add it as a package while bringing in the manifest, but I'm not sure how that's done. We need to add some things to [compat] in Project.toml I think to make things work smoothly.

from codingtheory.

Ah, I see. The offending line is Installed Oscar ──────────────────────────── v0.14.0. As Ben said, this is ignoring the Manifest file. I believe I have fixed it now by adding explicit version numbers to the Project file.

I didn't see why it would break in the function you specified, but the stacktrace shows it breaks in another function called by that, which makes sense. The Oscar v13 -> v14 update was a major breaking update. In this instance, v13 requires inputs to polynomial_ring in the form (:x, :y), but v14 now requires [:x, :y]. These changes are currently being made in the up_deps branch.

from codingtheory.

and to get GF you need using Oscar

from codingtheory.

The upshot is, @cyanreg, give Pkg.add(url="https://github.com/esabo/CodingTheory") another shot. It might work directly now without extra intervention.

from codingtheory.

Thanks, everything works fine now.

from codingtheory.