INSTALL-MATPLOTLIB.md

tinyGS-antenna-map

This is the antenna performance plotted from tinyGS reception data. See their repository.

Install

Having a good Python 3 setup is vital. Getting a good Matplotlib setup is tricky, but doable.

The basics - setting up Python 3 in a Raspberry Pi (or similar)

This code is Python 3 and the following system cleanup would be useful for many other systems/programs. This confirms that only Python 3 is installed and that it's cleanly running!

$ sudo apt-get update
...
$ sudo apt-get install -y python3-pip
...
$ sudo apt-get remove -y python2
...
$ sudo update-alternatives --install /usr/bin/python python /usr/bin/python3 10
...
$

Installing Matplotlib

After a hunk of testing, I've decided that using the system packages for Matplotlib (this will include numpy) is the best way to go.

$ sudo apt-get install -y python3-matplotlib
...
$

You can check which version you are running via:

$ dpkg-query -l python3-matplotlib python3-numpy
..
ii  python3-matplotlib 3.0.2-2      armhf        Python based plotting system in a style similar to Matlab (Python 3)
ii  python3-numpy      1:1.16.2-1   armhf        Fast array facility to the Python 3 language

$

or

$ python3 -c 'import matplotlib,numpy;print(matplotlib.__version__);print(numpy.__version__)'
3.0.2
1.16.2
$

Special notes on Matplotlib

This section is not normally needed; but provided for users that want newer libraries

The Matplotlib library requires the somewhat-matching numpy library. On a Raspberry Pi (and maybe other systems) it's best to install these packages via system commands (vs pip3). However, you can upgrade them to the latest code (if you really desire) using these commands:

$ sudo apt install -y python3-numpy libopenjp2-7-dev libtiff5 libatlas-base-dev
...
$ sudo python3 -m pip install -U matplotlib numpy
...
$

There's plenty of issues with this method of installation. More can be found via numpy issue 14772 and stackoverflow's vast collection of answers.

if you see the following error ...

RuntimeError: module compiled against API version 0xe but this version of numpy is 0xd

... it's a case of Matplotlib and numby being out of sync. Make sure both are updated via the pip3 command above. See numpy issue 655 and matplotlib issue 10135 and yet another stackpath answer etc etc.

Next steps

Please return to the README file.