pymwalib 0.8.3

A Python interface for mwalib, a library to read Murchison Widefield Array (MWA) raw visibilities, voltages and metadata into a common structure

pymwalib

A Python interface for MWATelescope/mwalib that provides a simple interface for reading MWA correlator raw visibilities, recombined voltages and exposing the relevant metadata. pymwalib and mwalib both support reading data produced by the existing (legacy) MWA and the new MWAX systems.

Prerequisites

• Python >=3.6
• mwalib (pymwalib release major version should match mwalib major version- for prerelease versions <1.0.0, compatibility is via the minor version number 0.N.*). See Installing mwalib section below.
• numpy

Installing mwalib

• Download the relevant release from MWATelescope/mwalib github releases

$ wget https://github.com/MWATelescope/mwalib/releases/download/v0.3.0/libmwalib-0.8.3-linux_x86_64.tar.gz -O libmwalib-0.8.3-linux_x86_64.tar.gz 

• Create a directory and extract the libmwalib-X.X.X-linux_x86_64.tar.gz tarball file. (Where X.X.X is the relevant version number).

$ mkdir -p mwalib
$ tar -xvf libmwalib-0.8.3-linux_x86_64.tar.gz -C mwalib

• The next step can be done one of two ways:
  • A) Copy the header and library files into your system /usr/local/include and /usr/local/lib paths.

$ sudo cp lib/mwalib.* /usr/local/lib/.
$ sudo cp include/mwalib.h /usr/local/include/.

• OR B) Append the path to the library files to your LD_LIBRARY path. /absolute/path/to/mwalib/lib/directory is the absolute path to the lib directory from the extracted mwalib release. NOTE: to make this change permanent for any new login sessions you should add the export line to your .bashrc file.

$ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/absolute/path/to/mwalib/lib/directory

Install pymwalib from pypi

You can install pymwalib via pip. Before doing so, you may want to create a new virtual environment, first. Then:

$ pip install pymwalib

OR Build pymwalib from source

Clone this repository

• Clone master or a specific release / commit- the example below will clone the master branch.

$ git clone https://github.com/MWATelescope/pymwalib.git pymwalib

Setup virtual environment and install requirements

• Create virtual environment

$ cd pymwalib
$ python3 -m venv env

• Activate the virtual environment

$ source env/bin/activate
(env)$ 

• Install requirements

(env)$ pip install -r requirements.txt 

Build pymwalib

• Run the setup.py to build the library

(env)$ python setup.py install

Test

• Run one of the example scripts against an MWA observation (to download MWA Observations, please visit the MWA ASVO data portal).

(env)$ python examples/print-context.py -m path_to_metafits.metafits path_to_gpuboxfile01.fits path_to_gpuboxfile02.fits ... 

Docker

Build the image

You can also build a docker container with pymwalib preinstalled:

$ docker build . -t pymwalib:latest

Run the container

One way to launch the container is to use docker run, for example:

docker run --name my_pymwalib --volume=/host_data:/data --entrypoint="" --rm=true pymwalib:latest python /pymwalib/examples/print-context.py -m /data/1101503312.metafits /data/1101503312_20141201210818_gpubox01_00.fits /data/1101503312_20141201210818_gpubox02_00.fits 

• This assumes /host_data is a directory on the machine running docker (host) containing the gpubox files and metafits for an observation.