Skip to content

An experimental visualization system for anonymous publish subscribe architectures.

License

Notifications You must be signed in to change notification settings

alexandrx/rqt_graphprofiler

 
 

Repository files navigation

rqt_graphprofiler

A ROS System Profiling and Visualization tool

Introduction

rqt_graphprofiler is a ROS profiling and visualization tool. This project was started by Dan Brooks (@mrdanbrooks) at @osrf/rqt_graphprofiler. The initial proposal for this tool is detailed in: ROS System Profiling and Visualization

Different from rqt_graph, which produces a rather simple display of nodes and their connections using dot, rqt_graphprofiler renders more intuitive graphs of nodes, published/subscribed topics, bandwidths, CPU and memory usage, etc.

Running it

rqt_graphprofiler needs the rosprofiler package to get the statistics data.

Running rosprofiler

rosprofiler needs the /enable_statistics parameter:

rosparam set enable_statistics true

Now run rosprofiler and rosgrapher on different terminals:

rosrun rosprofiler rosprofiler
rosrun rosprofiler rosgrapher

Running rqt_graphprofiler

rosrun rqt_graphprofiler rqt_graphprofiler

It allows to blacklist topics and nodes we don't need to see and also output SVG images.

Issues

rqt_graphprofiler has been without activity since 18 Oct. 2014, the latest release is 0.1.2. Many of the PyQt functions it uses have gone deprecated and doesn't work in newer releases of ROS. An effort was made to overcome such limitations and release 0.1.3 was proposed.

Todo

  • Support for PyQt5 and ROS Kinetic
  • Fix mouse zoom/drag issues
  • Added more information in nodes and bands
  • Added variable width of nodes and horizontal text layout
  • Evaluate nodes and machine location
  • Fix incorrect alignments of nodes' text
  • Add curved bands as shown in the original document
  • Improve bands sorting and widths
  • Fix spurious crashes
  • Support SVG output
  • (optional) Support PNG output
  • (optional) Support PDF output

About

An experimental visualization system for anonymous publish subscribe architectures.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 97.3%
  • Makefile 2.4%
  • CMake 0.3%