aardk / recipe

Minimial recomputation for explorative data analysis

RECIPE: Minimal recomputation for radio astronomy

This is version 2 of recipe which focuses on memoization. See the PASP article: Nikolic, Small & Kettenis (2017) "Minimal Re-computation for Exploratory Data Analysis in Astronomy". Licensed under GPL (like the dependency CASA).

How to install

export CASADIR=/home/user/p/casa-release-5.1.0-74.el7/
wget https://bootstrap.pypa.io/get-pip.py
${CASADIR:? must be set}/bin/python get-pip.py
${CASADIR}/bin/pip2.7 install --upgrade setuptools

Then either install from GitHub

${CASADIR:? must be set}/bin/pip install git+https://github.com/bnikolic/recipe.git

OR install from internet

${CASADIR:? must be set}/bin/pip install --extra-index-url=https://www.mrao.cam.ac.uk/~bn204/soft/py recipe

How to use

The two main things to remember about how Recipe works:

1. Output will always be contained in a directory returned by the function (assign it to a variable!). Any explictly specified output will be ignored.

2. If a task modified it's input, the modified input will be in the output directory

from recipe import casatasks as c
c.importuvfits(....)
c.clean(....) 

How to contribute

git clone http://www.mrao.cam.ac.uk/~bn204/g/recipe/
<work>
git send-email --to="[e-mail]"

Or if you prefer GitHub, see: https://github.com/bnikolic/recipe