Paper introducing jax-cosmo

Welcome to the jax-cosmo paper project. This paper is being written as a collaborative and open project.

Scope of the paper

Here is the current main scope and aims of the project:

• Introduce the jax-cosmo library
• Demonstrate to the cosmology community why differentiability is good with a few examples on a standard setting (e.g. DES Y1). Here are a few possibilities:
  • Forecast demo
  • Linear data compression demo a la MOPED
  • Fast inference with VI or HMC

But we will generally leave actual science investigations out of this paper, these can be explored in independent papers, outside of the scope of the Initiative or the jax-cosmo project. The idea being that you can contribute to this library, and still get a first author paper on a science application using that tool.

Guidelines for authorship

Here are the proposed authorship guidelines for this paper:

• We welcome any and all contributions, which will automatically grant you authorship. All contributions will be documented at the end of the paper.

• We will separate two tiers in authors:

  • First tier: for people actively writing the paper and/or contributing to the code and demos.
  • Second tier: for people helping with feedback, ideas, etc.
    Typically first tier will mean at least writing a section of the paper, adding a functionality to the code, or making a plot for the paper.

• First tier authors will appear sorted by amount of contribution, as first evaluated as fairly as possible by the instigator of the project (i.e. in this case @EiffL) and after consensus of first tier authors. Second tier will be ranked alphabetically.

These guidelines are an attempt at a fair contribution recognition model, please file an issue to propose modifications and/or raise any concerns.