Signal View plugin for PlotJuggler

An oscilloscope-style signal viewer plugin for PlotJuggler. Displays time series data as step-wise (staircase) traces with individually positionable Y-axis bars, multi-file overlay comparison, and time shifting.

Companion plugins:

• plotjuggler-zenoh — DataStreamer plugin for live data over Zenoh

Features

• Step-wise signal traces with per-pixel downsampling for smooth rendering at any zoom level
• Three-column Y-axis panel: signal names, cursor readout values, and draggable Y-axis bars with tick marks
• Multi-file data overlay: load CSV files on top of PlotJuggler data for side-by-side comparison; right-click the data-sets footer to remove individual overlays or clear all at once
• Time shifting: drag to shift any data layer's time offset for alignment
• Horizontal zoom: rubber-band drag or scroll wheel to zoom the time axis; right-click to pan (pre-pan range saved to history); Prev Zoom button steps back through the full zoom history
• Flexible signal bands: independently position, resize, and group signal display regions
• Signal context menu: right-click any signal's name, value, or Y-axis bar to auto-scale, change style, group, or remove — applies to all selected signals when the right-clicked signal is part of the selection
• Selection and batch editing: select signals with click, Ctrl+click, rubber band, or Ctrl+A; edit properties in bulk via the Y-range table dialog
• Customizable appearance: per-signal color, line style, line width, marker style, and Y-axis divisions
• Streaming mode: auto-scroll the time axis to follow live DataStreamer data with configurable buffer time; auto-enables on stream start, pauses data on user interaction, and resumes cleanly
• Layout persistence: full save/restore of signal configuration, overlays, streaming state, and view state via PlotJuggler layout files
• Plot cursor navigation: double-click to jump the cursor, drag within 5 px to slide it, or use left/right arrow keys for single-pixel steps
• Dark/Light theme: toggle between dark (default) and light mode; signal colors are converted via an invertible HSL lightness mirror so switching back always restores the original colors exactly; theme is saved and restored with the layout

Usage

1. Click Add Signal to add signals to the view
2. Double-click a signal name, value, or bar to edit properties (color, line style, Y range, etc.)
3. Right-click a signal name, value, or bar to auto-scale, change style, group, or remove (applies to all selected signals)
4. Click Overlay to load a CSV file for comparison; right-click the data-sets footer to rename, remove, or clear all overlays
5. Toggle Time Shift and drag on the canvas to align data layers
6. Toggle H. Zoom and drag to zoom into a time range, or scroll the wheel to zoom centered on the mouse; right-click to pan
7. Click Prev Zoom to step back through previous zoom ranges one at a time
8. Toggle Stream for live data: auto-scrolls with incoming data; pan/zoom/cursor drag pauses the stream; click Stream again to resume
9. Click Dark/Light to switch themes

Installing from Releases

Pre-built plugin binaries are available on the Releases page. Each release is built against a specific PlotJuggler version (noted in the filename and release notes).

1. Download the plugin file for your OS:

   • Linux: libSignalViewPlugin-<version>-pj<pj_version>.so
   • Windows: SignalViewPlugin-<version>-pj<pj_version>.dll

2. Copy the file into PlotJuggler's plugins directory:

   OS	Plugins directory
   Linux (source install)	<plotjuggler-install>/lib/plotjuggler/plugins/
   Linux (apt/ROS)	/opt/ros/<distro>/lib/plotjuggler/plugins/ or /usr/lib/plotjuggler/plugins/
   Windows	<plotjuggler-install>\lib\plotjuggler\plugins\

3. Restart PlotJuggler — Signal View will appear under Tools.

Building from Source

Requirements

• PlotJuggler 3.x (built and installed from source)
• Qt 5
• CMake 3.16+
• fmt library (bundled with PlotJuggler)

Linux (Ubuntu/Debian):

sudo apt install qtbase5-dev cmake

Windows:

• Visual Studio 2026 with the v142 (VS 2019) C++ toolchain installed
• Qt 5.15 via aqtinstall: pip install aqtinstall && aqt install-qt windows desktop 5.15.2 win64_msvc2019_64 -O C:\Qt
• PlotJuggler must be built from source and installed via cmake --install to produce the install tree

Building

Linux

If plotjuggler is on your PATH, the build script auto-detects the install prefix:

./build.sh

Otherwise, set PJ_INSTALL_DIR manually:

PJ_INSTALL_DIR=/path/to/plotjuggler-install ./build.sh

Windows

If both qmake and plotjuggler are on your PATH, the build script auto-detects both:

build.bat

Otherwise, set the directories manually:

set QT_DIR=C:\Qt\5.15.2\msvc2019_64
set PJ_INSTALL_DIR=C:\path\to\plotjuggler-install
build.bat

The script defaults to Visual Studio 2026. To use Visual Studio 2022 instead:

build.bat vs2022

Both generators require the v142 (VS 2019) C++ toolchain to be installed, as Qt 5.15 was built against it.

Running

Linux

./run.sh                        # launch PlotJuggler with the plugin
./run.sh -d data.csv            # load a data file
./run.sh -l layout.xml          # restore a saved layout

Windows

run.bat                         # launch PlotJuggler with the plugin
run.bat -d data.csv             # load a data file
run.bat -l layout.xml           # restore a saved layout

Any additional arguments are passed through to PlotJuggler.

Then go to Tools > Signal View to open the plugin.

Testing

Linux

Run the unit tests with:

./test.sh

This builds with Google Test and runs all test suites via CTest. Tests cover name parsing, overlay manager logic, axis layout calculations, CSV parsing, UI widget interactions, and Dark/Light theme logic.

To output JUnit XML results (used by CI for test reporting):

JUNIT_OUTPUT=test-results.xml ./test.sh

Windows

test.bat

To use Visual Studio 2022:

test.bat vs2022

To output JUnit XML results:

set JUNIT_OUTPUT=test-results.xml
test.bat vs2022

Coding Conventions

• Google C++ formatting (clang-format -style=Google)
• Qt naming conventions: m_ member prefix, UPPER_CASE constants, camelCase functions, PascalCase classes

Project Structure

src/
  signal_view_plugin.h/cpp   Plugin entry point (PJ::ToolboxPlugin)
  signal_view_widget.h/cpp   Main widget: toolbar, layout, signal management
  plot_canvas.h/cpp          Time series rendering and mouse interaction
  y_axis_panel.h/cpp         Signal name, value, and Y-axis bar columns
  overlay_manager.h/cpp      Multi-layer data management and CSV loading
  data_sets_panel.h/cpp      Footer panel listing loaded data layers