liveplot.py

liveplot is a system for minimal hassle, on-the-fly, dataset visualization in python. If you want maximum customizability, or plot types other than line plots and image plots, look elsewhere. But if you just want to see your data as it comes in to your script, with minimal effort, and without the possibility of graphics bugs crashing your script, this might be for you.

Liveplot works in two processes, one of which is a pyqt application hosting the window, the other is your script, which sends data to the window over a named pipe. The intended workflow is to open the window once (per session) and keep it open, rather than restarting it for every run of the script. Ideally, multiple scripts can communicate with the same window, and data remains available in the window until it is overwritten.

Requirements

• Numpy
• PyQt4
• pyqtgraph

pyqtgraph will be installed automatically from PyPI if not found

Basic Usage

Install from PyPI

pip install liveplot

or from the source directory

python setup.py install

and start the window

python -m liveplot

on windows, if you have py2exe, build the executable

python setup.py 

which should produce dist/liveplot.exe

If the window has been successfully started, open a client and plot

from liveplot import LivePlotClient
import numpy as np
plotter = LivePlotClient()
xs = np.linspace(0, 10, 100)
plotter.plot_xy('my test data', xs, np.sin(xs))

Individual plots are specified by their name, which can be any unique string. Attempting to create two different types of plot with the same name is currently an error. See more examples with the test suite,

python liveplot_test.py

Several methods of plotting are supported, including cumulative, parametric, and 2D-Image.

OS X Setup

By default Macs restrict the size of shared memory that can be allocated to a single process. To overcome this, copy sysctl.conf to /etc, or append it if the file is already present. Reboot your system to apply the changes.

GUI Features

In addition to the many wonderful features of native pyqtgraph widgets we have,

• Double click on plots to bring up cross-hair marker
• Cross-hair displays cross-section cuts for image plots
• Restore closed plots by double-clicking the name in the plot list
• Focus on a single plot by maximizing
• Right click on image plots
  • toggle histogram & levels scale
  • enable/disable auto-rescaling of levels when image is updated