Installation

Package dependencies

Mandatory dependencies:

  • tensorflow

  • numpy

  • scipy

  • pandas

  • h5py

  • tqdm

Optional dependencies:

  • refl1d: only if fast generation of reflectivity curves is necessary (e.g. for training)

  • matplotlib: only if you want to see plots

  • fabio: only for parsing of .fio files

Installation via pip

The package is mainly tested for Python 3.6, but there is a high probability that it might also work on later version. It can simply be installed from the well-known PyPI repository via the pip tool via:

python3.6 -m pip install mlreflect

This will by default install the latest version of mlreflect, which is usually preferable due to continued improvements, new features and bug fixes. However, please observe whether the the package version this document refers to is the same as the installed version since between versions, changes to the API might occur. For very recent versions there might not yet be a up-to-date documentation.