e7bc025379
Summary: the cred calculation is defined by a Markov Mixing process. By introducing the seed vector and teleportation parameter alpha, the Markov mixing process is augmented with a source of cred originating from the seed vector. The resulting algorithm is the generalized variation of PageRank, allowing computation of both canonical PageRank where the seed vector is the uniform distribution and personalized PageRank where the seed vector is an indicator distribution. It is still possible to get the simple markov chain solution by setting alpha = 0. Note that this changes the Markov process state update, but does not provide updates to the APIs. All existing behavior is unchanged because alpha is always set to 0. This is a port of https://github.com/sourcecred/odyssey-hackathon/pull/3, which was created during the Odyssey Hackathon. Test Plan: Existing tests have been extended to include passing alpha = 0 to reproduce exisiting test cases for the simple Markov Process. Additional test cases include - Verifying that resulting stationary distribution is unaffected by seed when alpha = 0 - Verifying that resulting stationary distribution is precisely equal to seed when alpha = 1 - Verifying that the resulting stationary distribution is linear in the seed vector - Verifying that the correct stationary distribution is computed for non-zero alpha - verify that the algorithm converges immediately when the initialDistribution is the stationary distribution - verify that the changing the initialDistribution does not change the stationary distribution Paired with @mzargham
SourceCred
SourceCred creates reputation networks for open-source projects. Any open-source project can create its own cred, which is a reputational metric showing how much credit contributors deserve for helping the project. To compute cred, we organize a project’s contributions into a graph, whose edges connect contributions to each other and to contributors. We then run PageRank on that graph.
To learn more about SourceCred’s vision and values, please check out our website and our forum. One good forum post to start with is A Gentle Introduction to Cred.
For an example of SourceCred in action, you can see SourceCred’s own prototype cred attribution.
Current Status
We have a prototype that can generate a cred attribution based on GitHub interactions (issues, pull requests, comments, references, etc.). We’re working on adding more information to the prototype, such as tracking modifications to individual files, source-code analysis, GitHub reactions, and more.
Running the Prototype
If you’d like to try it out, you can run a local copy of SourceCred as follows. First, make sure that you have the following dependencies:
- Install Node (tested on v8.x.x).
- Install Yarn (tested on v1.7.0).
- Create a GitHub API token. No special permissions are required.
- For macOS users: Ensure that your environment provides GNU coreutils. See this comment for details about what, how, and why.
Then, run the following commands to clone and build SourceCred:
git clone https://github.com/sourcecred/sourcecred.git
cd sourcecred
yarn install
yarn backend
export SOURCECRED_GITHUB_TOKEN=YOUR_GITHUB_TOKEN
node bin/sourcecred.js load REPO_OWNER/REPO_NAME
# this loads sourcecred data for a particular repository
# it can take a few mins to run and will exit when finished
yarn start
# then navigate to localhost:8080 in your browser
For example, if you wanted to look at cred for ipfs/js-ipfs, you could run:
$ export SOURCECRED_GITHUB_TOKEN=0000000000000000000000000000000000000000
$ node bin/sourcecred.js load ipfs/js-ipfs
replacing the big string of zeros with your actual token.
You can also combine data from multiple repositories into a single graph.
To do so, pass multiple repositories to the load command, and specify an “output name” for the repository.
For instance, the invocation
node bin/sourcecred.js load ipfs/js-ipfs ipfs/go-ipfs --output ipfs/meta-ipfs
will create a graph called ipfs/meta-ipfs in the cred explorer, containing the combined contents of the js-ipfs and go-ipfs repositories.
Early Adopters
We’re looking for projects who want to be early adopters of SourceCred! If you’re a maintainer of an open-source project and would like to start using SourceCred, please reach out to us on our Discord or our forum.
Contributing
We’d love to accept your contributions! You can reach out to us by posting on our forum, or chatting with us on Discord. We'd be happy to help you get started and show you around the codebase. Please also take a look at our contributing guide.
If you’re looking for a place to start, we’ve tagged some good first issues.
License
SourceCred is dual-licensed under Apache 2.0 and MIT terms:
- Apache License, Version 2.0, (LICENSE-APACHE or https://www.apache.org/licenses/LICENSE-2.0)
- MIT License (LICENSE-MIT or https://opensource.org/licenses/MIT)
Acknowledgements
We’d like to thank Protocol Labs for funding and support of SourceCred. We’d also like to thank the many open-source communities that produced the software that SourceCred is built on top of, such as Git and Node.