BARK
Getting StartedBuilding From SourceCore ModulesDeveloping Behavior ModelsBenchmarkingAdvanced TopicsBARK-ML
Getting StartedObserversEvaluatorsEnvironmentsBARK-ML AgentsThe following prerequisites are required to build BARK from source:
python-tk
.pip install virtualenv==16.7.8
First, clone the BARK repository and enter the folder: git clone https://github.com/bark-simulator/bark.git && cd bark
.
Create a virtual Python environment and install the dependencies of BARK by running bash install.sh
and enter it using source dev_into.sh
.
Before running any examples or implementing new code, it is advised to make sure all modules of BARK work properly by running all tests using the command bazel test //...
.
Finally, you are ready to implement your own code or to run the first example that has been built from source using, e.g., bazel run //bark/examples:merging
.
A detailed video for installing BARK is shown here:
The description from above also applies for building BARK from source on macOS. If you run into trouble with TKInter have a look here or the Matplotlib TKAgg backend have a look here.
Make sure that there is a 'Python.h' file in the bark/python_wrapper/venv
folder.
You might be running out of memory during the bazel build.
Try limiting the memory available to bark via
bazel build //... --local_ram_resources=HOST_RAM*.4
(or any other build or test call).