Generating node graphs from Finesse scripts

166  Sean Leavey  

There is a somewhat “hidden” feature in Pykat which is capable of showing you an approximate visualisation of your model. It may not be available on all systems, e.g. Windows.

This feature is available in two ways:

Via notebook

In a Jupyter notebook, you can view your interferometer using the following code:

from pykat.tools.plotting.graph import NodeGraph
nodegraph = NodeGraph(kat)
nodegraph.view_pdf()

In the example above, the kat object is your Pykat model. If you have called it something else, e.g. kat2, you should of course update this variable.

Via command line interface

You may not know this, but by installing Pykat you have also installed a command line interface. Within a Conda terminal/prompt, you can type pykat then enter, and you will be presented with a help screen for the command line interface. Normally you run a Finesse model this way using pykat FILE.kat where FILE.kat is your input file. Note that this interface does not support Python-based scripts, such as those that you write in your Jupyter notebooks. One option is --display-graph, which runs the model contained in the specified FILE.kat parameter and displays the graph in your PDF viewer.

Output

The output looks something like this (it will open a PDF on your computer):

The output from the node graph tool for a complicated Virgo model.

One comment on “Generating node graphs from Finesse scripts”

  1. By default, graphviz – the program we use to generate the node graph – is using a particular algorithm to lay out the components. An even better algorithm for plotting the node graph is called “neato”. You can use it by modifying the last line of code above:

    nodegraph.node_graph(engine='neato').view()

    This tends to put the beam splitter in Advanced LIGO into the middle of the graph, which is nice.

Leave a Reply