Skip to content

Latest commit

 

History

History
107 lines (94 loc) · 3.84 KB

installation.md

File metadata and controls

107 lines (94 loc) · 3.84 KB

Installation

We provide step-by-step instructions to install the nuPlan devkit. For a high-level overview, please read the general readme first.

Download the devkit

Download the devkit and move inside the folder:

cd && git clone https://github.com/motional/nuplan-devkit.git && cd nuplan-devkit

The above will download the files to your home directory. While you can change this to an arbitrary directory, the rest of our tutorials assumes that you are using the home directory.

Install Python

The devkit is tested for Python 3.9 on Ubuntu.

  • For Ubuntu: If the right Python version is not already installed on your system, install it by running: 
    sudo apt install python-pip
sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt-get update
sudo apt-get install python3.9
sudo apt-get install python3.9-dev
  • For Mac OS download and install from https://www.python.org/downloads/mac-osx/.

Install virtual environment

Next we setup a virtual environment. We recommend Conda for this purpose.

Install miniconda

See the official Miniconda page.

Create a Conda environment

We create a new Conda environment using environment.yml.

conda env create -f environment.yml

Activate the environment

If you are inside the Conda environment, your shell prompt should look like: (nuplan) user@computer:~$ Going forward, we shall always assume the Conda environment is active. If that is not the case, you can enable the virtual environment using:

conda activate nuplan

To deactivate the virtual environment, use:

conda deactivate

Install the devkit

Option A: Install PIP package from remote

Note: This option is not yet supported. It will be added soon.

For beginners, the easiest option is to install the PIP package:

pip install nuplan-devkit

This installs the devkit and all of its dependencies.

Option B: Install PIP package from local

We recommend that you install the local devkit as a PIP package:

pip install -e .

This installs the devkit and all of its dependencies. Note that the editable mode (-e) is optional and means that the code is used in-place, rather than being copied elsewhere, and can be modified for easy development.

Option C: Run source code directly

Alternatively, if you don't want to use the pip package, you can manually add the nuplan-devkit directory to your PYTHONPATH environmental variable, by adding the following to your ~/.bashrc:

export PYTHONPATH="${PYTHONPATH}:$HOME/nuplan-devkit"

To activate these changes you then need to run:

source ~/.bashrc

To install the dependencies, run the following command:

pip install -r requirements_torch.txt
pip install -r requirements.txt

Change default directories

As described in the general readme, the default nuPlan directories are:

~/nuplan/dataset    -   The dataset folder. Can be read-only.
~/nuplan/exp        -   The experiment and cache folder. Must have read and write access.

If you want to change these on your system, you need to set the corresponding environment variables in your ~/.bashrc, e.g.:

export NUPLAN_DATA_ROOT="$HOME/nuplan/dataset"
export NUPLAN_MAPS_ROOT="$HOME/nuplan/dataset/maps"
export NUPLAN_EXP_ROOT="$HOME/nuplan/exp"

This step is also required if you want to run any of the unit tests in the devkit.

That's it you should be good to go!