matisse-controller 0.3.0

A Python package to control the Matisse 2 TS.

matisse-controller

A Python package to control the Matisse 2 TS laser for the University of Washington's Optical Spintronics and Sensing Lab.

Requirements: Python 3.7+, NI VISA, PyVISA, pySerial, SciPy, matplotlib, PyQt5

Tested on Windows 10 (x64).

Installation

$ pip install matisse-controller

Usage

To launch the GUI, connect the Matisse and a supported wavemeter, and then run:

$ matisse-controller

To configure the behavior of the program, click the 'Configuration' menu option from the main GUI, or run:

$ matisse-config

The GUI uses a Python API to control the Matisse. If you're writing a Python program, just import the subpackages that contain the APIs you want. The matisse subpackage contains Matisse-related components, the config subpackage contains configuration functionality, etc.

Development

After checking out the repo, run pipenv install --dev to install dependencies. Using a virtual environment is recommended.

To install this package onto your local machine, run pip install -e ..

Useful documentation: PyVISA, pySerial, SciPy, matplotlib, Qt 5

Adding features to the Matisse class

The standard way of interacting with the Matisse outside of the existing API is to use the Matisse.query method. The Matisse implements several commands that run asynchronously, like motor movements, so if you want to run these synchronously, you must do it on your own (like checking the motor status until it's idle again).

Long-running tasks should be executed in a thread that can be started and gracefully stopped from the Matisse class. Currently, fetching a measurement from the wavemeter is a relatively expensive process, so avoid doing this too much if possible.

Adding another wavemeter

Currently I've only implemented an interface for the WaveMaster, but any class will do, as long as it implements the get_raw_value and get_wavelength methods. The get_raw_value method should return a value representing exactly what is seen on the wavemeter display (this might not be a measurement), and the get_wavelength method should always return a floating-point number representing the latest measurement from the wavemeter. The WaveMaster implementation blocks the thread until a value is returned from the instrument. Additionally, please ensure any code you write that communicates with instruments is thread-safe.

Adding features to the GUI

Logging and UI updates should have top priority, so take care not to block the UI thread. Here's the process I use:

• Add a menu action under setup_menus and connect it to a Qt slot under setup_slots, to be executed on the main thread later.
• Do UI updates in this slot, if you need a long-running task that also updates the UI, use a subclass of QThread with a slot (see LoggingThread, StatusUpdateThread for examples).
• For long-running tasks that do not need access to the UI, submit a runnable object to the ControlApplication's instance of ThreadPoolExecutor. Hold a reference to the Future it gives you and call add_done_callback on it, passing in ControlApplication.raise_error_from_future if you want to log errors from that thread. For an example of a method that runs tasks one-by-one on the Matisse, see ControlApplication.run_matisse_task.

Adding another PLE procedure

Currently I've only implemented a PLE scan for the Andor Shamrock 750. If you'd like to implement your own PLE procedure, create a separate Python package with a class that has the methods start_ple_scan, stop_ple_scan, and analyze_ple_data. It's up to you to implement the scanning logic for your particular spectrometer and CCD setup. Modify the Matisse class __init__ method to use your chosen wavemeter and an instance of your PLE scanning class.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/lucis-fluxum/matisse-controller.

License

The package is available as open source under the terms of the MIT License.