This GTlab module allows to run python scripts in intelligraph workflows.
- 🐍 Run python scripts as intelligraph nodes
- 🔌Custom input / output ports to pass python data
- 📊Matplotlib interface to draw plots directly in the node
The largest portion of the code is licensed under the Apache 2.0 License.
Smaller thirdparty party code part of code base uses other permissive licenses, such as the
BSD and MIT licenses. Please review the directory LICENSES and .reuse
for a full overview on all licensed used.
A working GTlab installation is required. This includes GTlab and the GTlab Logging library. In addition the python module and the intelligraph-module of the GTlab eco-system are required including thier dependencies.
| Library | Version | Bundled | Where to get |
|---|---|---|---|
| Qt | 5.15.x | No | https://download.qt.io/official_releases/qt/ |
| GTlab Core + Logging | >= 2.0 | No | https://github.com/dlr-gtlab/gtlab-core |
| Python | 3.9.x | No | https://www.python.org/downloads/ |
| PythonQt | >= 3.5.0 | Yes | https://github.com/MeVisLab/pythonqt |
| intelligraph-module | >= 0.12.0 | No | https://github.com/dlr-gtlab/intelligraph-module |
| python-module | >= v1.6.1 | No | https://github.com/dlr-gtlab/python-module |
The python node module requires a recent CMake (>3.15) to build. The configuration and build process is
similar to other CMake builds:
cmake -S . -B build_dir -DQt5_DIR=<path/to/cmake/Qt5> -DCMAKE_PREFIX_PATH=<path/to/gtlab_install> -DCMAKE_INSTALL_PREFIX=<path/to/gtlab_install>
cmake --build build_dir
cmake --build build_dir --target install
To display a plot in the node, there are a few things to consider:
- The script in the Python node has an interface for matplotlib. Only plots created with it can be displayed. The scripts must end with the show() function of the plot object.
- The node has a "Plot enabled" property. This must be set to true.
- The plot can only be displayed if the node has a certain size. If in doubt, you can check this by dragging the node larger (in the bottom right-hand corner of the node).
Connections of the nodes' ports can be used for data exchange between the nodes. Only ports of the same type can be connected to each other. The Python node can process all available types on the input side. Many provided data types have a value() function or similar to make the data of the connection accessible in the Python script. This is also demonstrated in the attached examples. A few generic data types are enabled on the output side. However, the intelli::ByteArrayData data type can be used to pass Python objects directly. In this case, the Python functionality "pickl" is used. This method is therefore restricted to data that is pickl-capable.
The data that is transmitted via the connections or ports can be called up in the script via their "Caption". This is set as a property in the corresponding window. This data object itself is not yet the value. This is usually obtained via a value() function or similar functions that can be implemented for the more specific data types. Output parameters can also be used directly and also have the caption set in the property system as the name in the script