Nifty tutorial for radio interferometric imaging

This repository provides a hands on tutorial to perform radio imaging using information field theory.

The demo scripts nifty_interfaces.py and nifty_intro.py give and introduction into the world of nifty and how to do inference with it.

In addition, two Jupyter notebooks are available:

• demo_CorrelatedFields.ipynb: An introduction to the correlated field model, its hyperparameters, and their effect in the statistical properties of the gaussian process.

• demo_radio.ipynb: A mock inference task given a simplified VLBI imaging setup using the uv-coverage of the 2017 imaging campaign of the Event Horizon Telescope (eht) and an artificially generated sky brightness distribution.

Requirements

• python3 (3.6 or later)

• matplotlib

• nifty8

• resolve

Installation via pip

To install the packages, simply run for example

pip install matplotlib
pip install nifty8
pip install ift-resolve

or build them from their sources.

Further information

To get started with the packages, take a look at the projects homepages:

• ift.pages.mpcdf.de/nifty/nifty8
• ift.pages.mpcdf.de/resolve

Additionally, the package nifty provides a variety of explanatory examples for imaging and general Bayesian inference tasks. See demos for more information.