Installation¶
Grond is compatible with Python 3.7 and above.
Manual installation (from source)¶
First install Grond’s prerequisites. It depends on
Next, get Grond’s source code using git clone:
git clone https://git.pyrocko.org/pyrocko/grond.git
cd grond
Finally, decide if you want to install Grond system-wide, for a single user or into a virtual or Conda environment.
System-wide installation:
# Assuming Pyrocko is already installed system-wide.
sudo pip3 install --no-deps .
Installation in your home directory:
# Assuming Pyrocko is installed system-wide or also in user home.
pip3 install --no-deps --user .
Installation into a Conda environment:
# Assuming Conda environment is active and Pyrocko has been installed, for
# example with `conda install -c pyrocko pyrocko`.
pip3 install --no-deps .
Installation into a virtual enviroment (venv):
# Assuming venv is active
pip3 install .
Note
The --no-deps option prevents pip from installing Grond’s
requirements automatically. Pip’s automatic dependency handling, only makes
sense in pure pip environments!
Updating a manual (from source) installation¶
For updating an existing manual installation of Grond, update the source code using git pull:
cd grond # change to the directory to where you cloned Grond initially
git pull origin master
Then reinstall as described in section Manual installation (from source) above.
Installation under Conda (Grond latest release package, from PyPI)¶
Pyrocko’s pre-built Conda packages are available online, Grond can be installed with Conda’s pip into the Conda environment.
conda install -c pyrocko pyrocko
pip3 install --no-deps grond
Installation with pip (Grond latest release package, from PyPI)¶
Grond and all its dependencies can be installed by running
pip3 install grond # read below...
but, we recommend to make a conscious decision about how its main dependency Pyrocko and especially Pyrocko’s own dependencies are installed. The Pyrocko Installation Manual describes different installation schemes.
As a general advice, we recommend to exclusively use either, (1) the system’s native package manager, (2) Conda, or (3) pip only. In (1) and (2), only resort to use pip for those few packages which are not available as native packages. Otherwise, competing package managers will ruin your day!
To prevent pip from automatically resolving dependencies run
pip3 install --no-deps grond
This assumes that Pyrocko and Kite have been installed beforehand.