The installation of PyScaffold only requires a recent version of of setuptools,
(at least version 46.1), pip, as well as a working installation of Git
(meaning at least your name and email were configured but also
setting the default branch might be useful in your first-time git setup).
Especially Windows users should make sure that the command
git is available on
the command line. Otherwise, check and update your
variable or run PyScaffold from the Git Bash.
pip install --upgrade pyscaffold
to get the latest stable version. The most recent development version can be installed with:
pip install --pre --upgrade pyscaffold
pip also has the advantage that all requirements are automatically
If you want to install PyScaffold with all official extensions, run:
pip install --upgrade pyscaffold[all]
conda install -c conda-forge pyscaffold
It is also very common for developers to have more then one Python version
installed on their machines, and a plethora of virtual environments spread all
over the place… Instead of constantly re-installing PyScaffold in each one of
these installations and virtual environments, you can use pipx to do a
“minimally-invasive” system-wide installation and have the
always available independently of which Python you are using:
pipx install pyscaffold
Please check the documentation of each tool to understand how they work with
extra requirements (e.g.
[all]) or how to add extensions (e.g.
inject pyscaffold pyscaffoldext-dsproject).
We strongly recommend installing tox together with PyScaffold (both can be installed
with pip, conda or pipx), so you can take advantage of its automation
capabilities and avoid having to install dependencies/requirements manually.
If you do that, just by running the commands
tox -e docs, you
should able to run your tests or build your docs out of the box (a list with
all the available tasks is obtained via the
tox -av command).
If you dislike tox, or are having problems with it, you can run commands (like
make -C docs) manually within your project, but then you
will have to deal with additional requirements and dependencies yourself.
It might be the case you are already have them installed but
this can be confusing because these packages won’t be available to other
packages when you use a virtual environment. If that is the case,
just install following packages inside the environment you are using for
If you have problems using PyScaffold, please make sure you are using Python 3.6 or greater.
In some operating systems, e.g. Ubuntu, this means installing a
python3-pippackage or similar via the OS’s global package manager.
conda is a very competent package manager for Python, not only when you have to deal with numbers. In general, when you rely on native extensions, hardware acceleration or lower level programming languages integration (such as C or C++), conda might just be the tool you are looking for.