PlotJuggler is a graphic interface that allows users to visualize data (and in particular time series) from either a rosbag or a streaming of messages from ros topics (including rosout).
It can be considered a unified replacement for rqt_plot, rqt_rosbag and rqt_console.
You can watch a short presentation video here:
The 1.0 release is a big milestone for the project. This is what you should expect from this first official release:
Feature-complete: there are a couple of new cool series on Netflix, so don’t expect me to do any further development in the near future. I will still fix bugs, though.
Reliability: the app apparently crushes less often than usual and I believe that was a compelling reason to tag the 1.0 release.
Stable API (plugins): … until I decide unilaterally to change it, of course.
In short, if you think that tagging the current release as “1.0” is just a marketing strategy to make you think that this software worth your trust, you are absolutely right.
I would like to thank all the user that relentlessly asked me to fix bugs and add features (without contributing with any PR, of course); with their Github issues they helped make PlotJuggler great again!
Cheers
Davide
P.S.= Seriously now. I appreciated a lot the people which sent suggestions, issue reports, feedback and PR.
If you find PlotJuggler useful, or think that it can be improved, please consider dropping me a line.
Very nice work! It really looks like a modernized rqt_plot! I will thoroughly look into it when I have the time. Watched, and starred on github for now! Keep it up!
Perhaps you might want to look at data formats commonly supported by other time series libraries, like pandas. I’m not sure how much demand there would be for JSON, HDF5, SQL, etc… but you might want to check: http://pandas.pydata.org/pandas-docs/stable/io.html
I was thinking about SQL and HDF5, but I am not sure about the number of people using these formats, and I am afraid to spend time on features no one really need
If anyone DOES need a specific format to be supported, let me know on Github issues.
I’ve started to see Line-Delimited JSON, especially in log streams and files. I have found it fairly useful in data logging streaming and log-analysis. I do not know how wide-spread it is in the ROS community currently.
The tricky part of HDF5, from what I can understand, is how to translate this generic fileformat into timeseries. If you want to get in touch to describe your particular use case, let me know.