Download and install Miniconda using the following instructions https://conda.io/projects/conda/en/latest/user-guide/install/windows.html
Open the Anaconda Prompt from the Start Menu to execute the following commands.
Add the conda-forge channel
conda config --add channels conda-forge
The conda-forge channel provides multitude of community maintained packages. Find out more about it here https://conda-forge.org/
Create a virtual environment in conda
conda create --name cubeenv python=3.6 datacube
Activate the virtual environment
source activate cubeenv
Find out more about managing virtual environments here https://conda.io/docs/using/envs.html
Install other packages
conda install jupyter matplotlib scipy
Find out more about managing packages here https://conda.io/docs/using/pkgs.html
Datacube is now installed and can be used in the Anaconda Prompt by activating the cubeenv environment.
Manual Installation (Fallback)¶
Instructions given in this section have not been updated or tested recently. We strongly recommend using Conda based installation on Windows platforms.
Only follow these steps if the Miniconda installation does not suit your needs.
Python 3 environment¶
If in a restricted environment with no local administrator access, python can be installed by running:
msiexec /a python-3.6.4.msi TARGETDIR=C:\Python3
Or by launching the version 3.6 installer and selecting not to install for all users (only single user install).
Ensure pip is installed:
cd C:\Python3 python -m ensurepip
Upgrade and Install python virtualenv:
python -m pip install --upgrade pip setuptools virtualenv
Create an AGDC virtualenv:
mkdir C:\envs Scripts\virtualenv C:\envs\agdcv2
3.5 only workaround: Copy
vcruntime140.dll from Python install dir into
The python virtual environment isolates this python installation from other python installations (which may be in use for other application software) to prevent conflicts between different python module versions.
On windows systems by default there are no ready configured compilers, and so libraries needed for some python modules must be obtained in precompiled (binary) form.
Download and install binary wheels from https://www.lfd.uci.edu/~gohlke/pythonlibs/
You will need to download at least:
The following may also be useful:
Install these packages by running in your
pip install *.whl
It may be necessary to manually replace
*.whl with the full filenames for each
.whl file (unless using a unix-like shell instead of the standard windows command line
For 3.5 only
If there are problems loading libraries. Try:
copy site-packages/matplotlib/msvcp140.dll site-packages/osgeo/
Also, install the python notebook interface for working with datacube example notebooks:
pip install jupyter
Obtain a current copy of the datacube source code from GitHub. A simple way is to extract https://github.com/opendatacube/datacube-core/archive/develop.zip into a subdirectory of the python environment.
Install the datacube module by running:
cd datacube-core-develop python setup.py install
Extra instructions for installing Compliance Checker¶
pip install cf_units
Download and install udunits2 from gohlke
Edit site-packages/cf_units/etc/site.cfg with path to udunits2.dll which should be venv/share/udunits/udunits2.dll