Skip to content

Home

Jump to bottom
elisachisari edited this page Jun 14, 2018 · 18 revisions

Welcome to the CCL wiki!

CCL is the Core Cosmology Library of LSST-DESC. This library provides us with cosmological predictions that we will be using to analyze the data coming from LSST. For example, for a given set of cosmological paramaters and survey properties, CCL can produce angular power spectra and correlation functions for clustering, lensing, etc.

How to get started with CCL

Clone the repository into your computer. Do

cd path/to/CCL/dir
git clone git@github.com:DarkEnergyScienceCollaboration/CCL.git

In order to compile CCL you need GSL version 2-1. You can get GSL here: https://www.gnu.org/software/gsl/. To install it you need to be the admin of your computer. Also in the future, you will also need FFTW3, which you can get here: http://www.fftw.org.

To compile CCL, go to the directory where you installed it and follow the instructions on the README. See also Known installation issues and our extensive README, which tells you how to customize the installation. Notice that the suggested installation procedure is different if you are planning on doing development for CCL.

How to contribute to CCL

We ask you that when you work on a modification on CCL, you do so by creating a git "branch". You can do this by going to the CCL directory in your computer and typing

git checkout -b nameofyourbranch

This command will at the same time create the branch and switch to it.

Now you can start your modifications. Once you are done and want to show us what you have done, you should do a commit

git commit -m "some message about what you have done"

and then push into the repository

git push -u origin nameofyourbranch

Continue working on your branch and push your changes into the upstream branch. To only push changes from your current branch, you need to configure git as follows

git config --global push.default current

When you are done with your work, open a pull request.

For more details, including how to work with the Python wrapper, please see our CONTRIBUTE file in the repository.

If you find something we have missed, please contribute by opening a github issue on this.

Relation to CosmoLike/Cosmosis

CosmoSIS: sampler, pipeline code, and collection of cosmology modules for inference with focus on flexibility

CosmoLike: optimized pipeline for fast, high-precision multiprobe predictions with focus on internal consistency

CCL is external to both these frameworks

  • Learn from strengths and weaknesses of both
  • Not replacing either of them, but can become part of each
  • Lives in its own repository
  • Designed to be independently useful however your team decides to work
Clone this wiki locally