Installation

Installation with pip or conda is currently neither supported nor necessary. Just clone the code repository from git:

git clone https://github.com/ncgeib/pypret.git

and the directory pypret within contains all the required code of the package. Either add its location to your PYTHONPATH or copy it in your working directory.

As the package matures I may add an installer.

Requirements

It requires Python >=3.6 and recent versions of NumPy and SciPy. Furthermore, it requires h5py for storage and loading. Optional dependencies are

• pyfftw (for faster FFTs)
• numba (for optimization of some low-level routines)
• python-magic (to recognize zipped HDF5 files)