Skip to content

Latest commit

 

History

History
153 lines (129 loc) · 6.23 KB

INSTALL_CONTRIBUTE.md

File metadata and controls

153 lines (129 loc) · 6.23 KB
Raw

refnx - Installation and Development Instructions

refnx is a python package for analysis of neutron and X-ray reflectometry data. It can also be used as a generalised curvefitting tool. It uses Markov Chain Monte Carlo to obtain posterior distributions for curvefitting problems.

Installation

refnx has been tested on Python 3.9, 3.10, 3.11, and 3.12. It requires the numpy, scipy, cython packages to work. Additional features require the pytest, pandas, qtpy, pyqt6, h5py, xlrd, attrs, tqdm, matplotlib, pymc, pytensor packages. To build the bleeding edge code you will need to have access to a C-compiler to build a couple of Python extensions. C-compilers should be installed on Linux. On OSX you will need to install Xcode and the command line tools. On Windows you will need to install the correct Visual Studio compiler for your Python version.

In the current version of refnx the emcee and ptemcee packages are vendored by refnx. That is, refnx possesses it's own private copy of the package, and there is no need to those packages separately.

Installation into a conda environment

Perhaps the easiest way to create a scientific computing environment is to use the miniconda package manager. Once conda has been installed the first step is to create a conda environment.

Creating a conda environment

  1. In a shell window create a conda environment and install the dependencies. The -n flag indicates that the environment is called refnx.

conda create -n refnx python=3.7 numpy scipy cython pandas h5py xlrd pytest tqdm attrs

  1. Activate the environment that we're going to be working in:
# on OSX
conda activate refnx

# on windows  
conda activate refnx

Installing into a conda environment from source

The latest source code can be obtained from either PyPi or Github. You can also build the package from within the refnx git repository (see later in this document).

  1. In a shell window navigate into the source directory and build the package. If you are on Windows you'll need to start a Visual Studio command window.
python setup.py build
python setup.py install
  1. Run the tests, they should all work.
python setup.py test

Installing into a conda environment from a released version

  1. There are pre-built versions on conda-forge, but they're not necessarily at the bleeding edge:

conda install -c conda-forge refnx 2) Start up a Python interpreter and make sure the tests run:

>>> import refnx
>>> refnx.test()

Development Workflow

These instructions outline the workflow for contributing to refnx development. The refnx community welcomes all contributions that will improve the package. The following instructions are based on use of a command line git client. Git is a distributed version control program. An example of how to contribute to the numpy project is a useful reference.

Setting up a local git repository

  1. Create an account on github.
  2. On the refnx github page fork the refnx repository to your own github account. Forking means that now you have your own personal repository of the refnx code.
  3. Now we will make a local copy of your personal repository on your local machine:
# <username> is your github username
git clone https://github.com/<username>/refnx.git
  1. Add the refnx remote repository, we're going to refer to the remote with the upstream name:
git remote add upstream https://github.com/refnx/refnx.git
  1. List the remote repositories that your local repository knows about:
git remote -v

Keeping your local and remote repositories up to date

The main refnx repository may be a lot more advanced than your fork, or your local copy, of the git repository.

  1. To update your repositories you need to fetch the changes from the main refnx repository:
git fetch upstream
  1. Now update the local branch you're on by rebasing against the refnx main branch:
git rebase upstream/main
  1. Push your updated local branch to the remote fork on github. You have to specify the remote branch you're pushing to. Here we push to the main branch:
git push origin main

Adding a feature

The git repository is automatically on the main branch to start with. However, when developing features that you'd like to contribute to the refnx project you'll need to do it on a feature branch.

  1. Create a feature branch and check it out:
git branch my_feature_name
git checkout my_feature_name
  1. Once you're happy with the changes you've made you should check that the tests still work:
python setup.py test
  1. If the performance of what you've added/changed may be critical, then consider writing a benchmark. The benchmarks use the asv package and are run as:
cd benchmarks
pip install asv
asv run
asv publish
asv preview

For an example benchmark look at one of the files in the benchmarks directory. 4) Now commit the changes. You'll have to supply a commit message that outlines the changes you made. The commit message should follow the [numpy guidelines][numpy-contib]

git commit -a
  1. Now you need to push those changes on the my_feature_branch branch to your fork of the refnx repository on github:
git push origin my_feature_branch
  1. On the main refnx repository you should be able to create a pull request (PR). The PR says that you'd like the refnx project to include the changes you made.
  2. Once the automated tests have passed, and the refnx maintainers are happy with the changes you've made then the PR is merged. You can then delete the feature branch on github, and delete your local feature branch:
git branch -D my_feature_branch