Install Poppy softwares

Warning If you want to install the board of a tangible robot, go to the startup chapter instead.

This section will guide you to install Poppy softwares on your personal computer. It is useful only if you are in one of these situations:

  • You want to control a simulated robot.
  • You want to control a Poppy creature from your computer without using the embedded board (Odroid or Raspberry Pi).

Poppy creatures are run by Python computer code. Depending on your operating system you will have to install Python and in any case you'll have to install the required software libraries. If you are getting started with Python and want to install a full Python environment for scientific computing, we suggest you to use Anaconda Python distribution.

Info Poppy software libraries works in Python 2.7 and Python 3.3+. However as we develop them in Python 2.7, so it is less likely bugs for this Python.

Table of content:

Install Python and Poppy software on Windows

If you want a step by step screencast of the installation of Anaconda on Windows, you can see these videos (this is a YouTube playlist).

Install Python

We encourage the use of the Anaconda Python distribution. However, if you already installed a Python distribution like Canopy (shipped with scientific packages), you can directly install Poppy software.

Anaconda

Download Anaconda Python distribution (400 MB) here for 64-bit computer or here for 32-bit.

Install it by clicking on "next" at each step. If you intend to install Anaconda for all users of your computer, be sure to select "all users".

Anaconda all users.

It is also very important that the two check-boxes of the PATH and the default Python are checked.

Anaconda install

Now you have a Python distribution ready to install Poppy software.

Miniconda (alternative to Anaconda)

Miniconda is a "light" version of Anaconda which contain only Python and the conda package manager. You can install it instead of Anaconda and save a lot of disk space (25 Mo vs 400 Mo), but you will have to do another step in the install process, and you will not have Jupyter notebook shortcut on the desktop. Download miniconda here for 64-bit computer or here for 32-bit computer.

Install it and be sure that the two check-boxes of the PATH and the default Python are checked.

Open the Command Prompt (press the windows key and type "Command Prompt"), type and press Enter to execute the command below:

conda install numpy scipy notebook jupyter matplotlib

Now you have a Python distribution ready to install Poppy software.

Install Poppy software

Open the prompt of your Python Distribution (called Anaconda Prompt for Anaconda) or the Command Prompt of Windows, type and press Enter to execute the command below: Anaconda all users.

pip install poppy-torso --user -U

This will install everything necessary to control a Poppy Humanoid. Substitute "poppy-torso" with "poppy-humanoid" or "poppy-ergo-jr" to install respectively a Poppy Humanoid or a Poppy Ergo Jr.

In case of update, it is advised to upgrade pypot (the motor library control) and the creature package separately:

pip install pypot --user -U --no-deps
pip install poppy-torso --user -U --no-deps

Install Python and Poppy software on Mac OSX

Mac OSX has a Python distribution installed by default. Before installing Poppy software, you need to install the Python package manager pip. Open a terminal copy and press enter to execute the command below:

curl --silent --show-error --retry 5 https://bootstrap.pypa.io/get-pip.py | sudo python

You can now install Poppy software for the creature of your choice:

pip install poppy-torso -U

Substitute "poppy-torso" with "poppy-humanoid" or "poppy-ergo-jr" to install respectively a Poppy Humanoid or a Poppy Ergo Jr.

Install Python and Poppy software on GNU/Linux

Most of GNU/Linux distributions, have already a Python distribution installed by default.

Using the default Python distribution

Pypot, the main library of the robot is depending (amongst some other) on two big scientific libraries Numpy and Scipy which are themselves depending on C and Fortran code. These libraries may be installed with the Python package system (pip), but because of the huge number and differences between GNU/Linux distributions pip is not able to distribute binaries for Linux so all dependencies must be compiled... The solution to avoid the compilation of numpy and scipy is to install them with your distribution package manager.

On Ubuntu & Debian:

curl --silent --show-error --retry 5 https://bootstrap.pypa.io/get-pip.py | sudo python
sudo apt-get install python-numpy python-scipy python-matplotlib python-dev

On Fedora:

curl --silent --show-error --retry 5 https://bootstrap.pypa.io/get-pip.py | sudo python
sudo yum install numpy scipy python-matplotlib

On Arch Linux:

curl --silent --show-error --retry 5 https://bootstrap.pypa.io/get-pip.py | sudo python
sudo pacman -S python2-scipy python2-numpy python2-matplotlib

You can now install Poppy software.

Note The downside is the Python libraries from you distribution system are very often out of date.

Using Anaconda (or miniconda)

If you want to have up to date numpy, scipy and ipython without having to compile them, we suggest you to install Anaconda or at least the conda package manager distributed with miniconda. Download miniconda (64-bit) with these command below in your terminal:

curl -o miniconda.sh http://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh

If you have a 32-bit computer

 curl -o miniconda.sh http://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86.sh

Execute commands below and follow the instructions to install miniconda:

chmod +x miniconda.sh
./miniconda.sh

You can now install some required and other useful dependencies for Poppy software with conda:

conda install numpy scipy notebook jupyter matplotlib

You can now install Poppy software.

results matching ""

    No results matching ""