Mantid MSlice

Build status

Overview

MSlice is a tool for performing slices and cuts of multi-dimensional data produced by Mantid.

Documentation

The current MSlice documentation can be viewed at

http://mantidproject.github.io/mslice

Development

The following setup steps are required regardless of the environment:

• install mantid from either http://download.mantidproject.org or yum/apt repositories (nightly on Linux)
• clone this repository

Command Line

To develop purely on the command line then simply use your favourite editor and run either

• mslicedevel.bat (Windows) or
• ./mslicedevel.sh (Linux)

Please note that you may have to update the path to your Mantid installation in the file first if you are not using a Mantid conda environment. For mslicedevel.bat one option is to set the path to your conda installation in 'CONDAPATH' and then simply run the batch file by double-clicking on it.

PyCharm

Mantid must be installed prior to running this setup.

To set up the PyCharm IDE first open PyCharm and select File->Open Project. Select the cloned mslice directory and select open. The project layout should be displayed. The first run may take some time to open while PyCharm parses the structure.

You will also need to edit the run configurations:

• The startup script is start_mslice.py.
• The bin directory of an installed version of Mantid must be on the PATH.
• Set the environment variable QT_QPA_PLATFORM_PLUGIN_PATH to the directory with the QT plugins from the Mantid installation MantidInstall/plugins/qt5.
• If you're developing on Windows, the Python interpreter used must be the one shipped with the Mantid installation MantidInstall/bin/python.exe.
• If you're developing on Ubuntu, set the Python Interpreter path to /usr/bin/python3.6

You can now also develop MSlice using a Mantid conda environment. First install Mantid using conda env create -f mslice-developer.yml, then add this interpreter by going to the File->Settings in PyCharm, then Project: mslice -> Python Interpreter, click the cog on the right side to add an existing interpreter and select Conda and Python 3.11 (mantidnightly). Then go to Run -> Edit Configurations and create new configuration with this interpreter. Specify start_mslice.py as the startup script.

If mslice doesn't launch, you may have to mark src folder as Sources Root on Pycharm

Optionally, you can also install pre-commit locally to ensure formatting issues are resolved when committing:

pre-commit install

Running tests

To run the unit tests in the MSlice repository, you can run them from PyCharm or from the command line.

To run them from PyCharm, create a pytest configuration and specify the Target as Custom. Enter the following configuration:

• Additional Arguments can be left empty to run all tests, or you can specify a test name.
• Python Interpreter should be your mantidnightly python executable
• Working directory should be the root of your Mslice repository

To run them from the command line, make sure you have your conda developer environment active, and cd to the root of this repository. You can then run a test using the following command:

pytest <test_name>

To check the coverage of all the tests in MSlice, you can run the following commands:

coverage run -m pytest
coverage report

Using Mslice on a Mantid Development Environment

If you are developing for Mantid on a local environment, mantid and mantidqt are installed as local packages, if you want to install the latest nightly mslice as a conda package, use the command: mamba install -c mantid/label/nightly mslice --no-deps on your mantid-developer environment. This avoids installing Mantid as a dependency of the mslice package.

Aditionally, a local mslice branch can be launched from mantid by setting the system path of the mslice source folder as an environment variable before starting mantid.

• On Clion/PyCharm run/debug configurations, set the PYTHONPATH on the Environment Variables line edit as: PYTHONPATH=/path/to/mslice/src/:$PYTHONPATH$
• On VSCode launch configurations in launch.json, set the field "env":{"PYTHONPATH":"/path/to/mslice/src/:$PYTHONPATH$"}" for a Python launch/debug configuration, or the field "environment":[ {"name":"PYTHONPATH", "value":"/path/to/mslice/src/:$PYTHONPATH$"}], for a C++ launch/debug configuration.

Nightly conda build

Every night the MSlice unit tests are run automatically using the latest nightly conda packages for mantid and mantidqt, as well as the main branch of MSlice. If the unit tests run successfully, and if changes have been made to the main MSlice branch within the last 24 hours, a new MSlice conda package labelled nightly is created and uploaded to the mantid conda channel.