This is the MIT version of the PyMca XRF Toolkit. Please read the LICENSE file for details.

Ready-to-use packages are available for the most common platforms.

PyMca frozen binaries for MacOS and Windows can be obtained from Sourceforge

Official packages are available for common Linux distributions.

Please continue reading if you want to use PyMca with your existing Python installation.

The simplest solution is to use pip:

pip install PyMca5

You can add the usual --user qualifier to install only for your local user rather than system-wide:

pip install PyMca5 --user

If you want to build from the source distribution or from a git repository checkout, you may want to have Cython installed on your system.

1. In your default system-wide python installation, run one or the other of the two (not both) commands below (may require root/administrator access):

   # Run one of the following (not both); pip is preferred
   python setup.py install    # use python setuptools
   pip install .              # use the pip package manager

2. Or, to install just in your local user account:

   # Run one of the following (not both); pip is preferred
   python setup.py install --user   # use python setuptool
   pip install . --user               # use the pip package manager

You will need the following dependencies installed:

• python (3.8 or higher recommended)
• numpy
• fisx
• h5py

If you want to use the graphical interface provided, you will need a running python installation with one of the following combinations:

• PyQt5 + matplotlib (PyMca license will be GPL unless you have a commercial PyQt5 license)
• PyQt6 + matplotlib (PyMca license will be GPL unless you have a commercial PyQt6 license)
• PySide6 + matplotlib (PyMca license will be MIT because PySide6 is LGPL)

If you want to embed PyMca in your own graphical applications, I recommend you to use the McaAdvancedFit.py module. It is very easy to embed.

To run the tests after installation run:

python -m PyMca5.tests.TestAll

• Use the fisx library for all Physics calculations and not just for corrections.
• Compound fitting.

If you have any questions or comments (or contributions!), please feel free to contact me or submit a pull request.

Enjoy,

V. Armando Sole