Programmatic usage

Shell

The Shell tab is a functional Python command line embedded within the program. The pipeline can be accessed directly from the shell, and behaves like a dictionary keyed by variable names. Pylab is imported in the shell making a number of MATLAB-style plotting and basic numeric commands accessible (see the matplotlib webpage for more docs). One can, for example, plot a histogram of point amplitudes by executing hist(pipeline['A']). Pipeline data sources can be accessed by entering pipeline.dataSources[datasource_key]. For a list of datasource keys, type pipeline.dataSources.

Jupyter notebook

PYMEVisualize can be used directly from a Jupyter notebook. At the top of the notebook, enter from PYME.LMVis import VisGUI, %gui wx, and then pymevis = VisGUI.ipython_pymevisualize(). This makes it possible to access a PYMEVisualize instance from a notebook through the pymevis variable. Setting pipeline=pymevis.pipeline gives the user access to the PYMEVisualize pipeline in exactly the same way as described in Shell section. An example of generating a point cloud, passing it to a tabular data source, and visualizing the data source is shown in Fig. 16.

Fig. 16 An example of generating a point cloud, passing it to a tabular data source, and visualizing the data source in PYMEVisualize from a Jupyter notebook.

Please note that both PYME and the Jupyter kernel must be set up in the Framework build on a Mac (see https://python-microscopy.org/doc/Installation/InstallationFromSource.html) for this to work. To install the Jupyter kernel in the Framework build, activate your_pyme_environment and then type PATH/TO/CONDA/ENVIRONMENT/python.app/Contents/MacOS/python -m ipykernel install --user --name your_pyme_environment.

Plugins

Details on extending and writing plugins for PYMEVisualize are available at http://python-microscopy.org/doc/hacking.html. A template for extending PYMEVisualize can be found at https://github.com/python-microscopy/pyme-plugin.