Installation#

Is cf-plot already installed?#

cf-plot my already be avaiable to you, if you have access to a some tier of HPC system. For example (though note these commands may be out of date and will be checked for updates in the near future):

On JASMIN#

$ export PATH=/home/users/ajh/anaconda3/bin:$PATH
$ ln -s /home/users/ajh/cfplot_data ~

On ARCHER#

$ export PATH=/home/n02/n02/ajh/anaconda3/bin:$PATH
$ export QT_XCB_NO_XI2=true
$ ln -s /home/n02/n02/ajh/cfplot_data ~

On Reading Academic Computing Cluster (RACC)#

$ module load ncas_anaconda3
$ ln -s /share/apps/NCAS/cfplot_data ~

To install cf-plot#

Linux and Mac OSX#

cf-plot is supported for Linux and Mac. There are multiple ways to install from such systems, as detailed below.

Via conda#

To install cf-plot on a Linux PC or Mac, get access to the conda command through an Anaconda or 'miniconda' distribution (etc.) and run:

Installing cf-plot and its dependencies with conda#

$ conda install -c conda-forge cf-python cf-plot udunits2
$ conda install -c conda-forge "esmpy>=8.7.0"  # if you need to use regridding functionality

The first line installs cf-python and cf-plot (udunits is a dependency of cf-python). The second line is optional and installs esmpy, together which cf-python uses for regridding data.

Via PyPI#

You can use pip to install from PyPI:

Installing cf-plot and its dependencies with pip#

$ pip install cf-python cf-plot

If you are upgrading the version of cf-python or cf-plot to the latest ones then add the --upgrade after the install above. A specific version can be installed using pip install cf-plot==3.3.0 for example.

From source via GitHub#

Installing cf-plot from source via git and GitHub#

$ git clone https://github.com/NCAS-CMS/cf-plot.git
$ cd cf-plot
$ python setup.py install  # or 'pip install -e .'

Note you will need to ensure all dependencies are avialable, such as cf-python. See below for guidance on this.

Windows#

Only Linux and Mac operating systems are directly supported, but if you wish to use Windows there are a couple of options whereby you can run Linux from Windows and use cf-plot that way:

Install the Microsoft Windows Subsystem for Linux (WSL). Once this is working install cf-python and cf-plot as per the Linux instructions above.
Installing a Linux Virtual Machine. Installation instructions and a Mint Linux Virtual Machine are available at http://gws-access.ceda.ac.uk/public/ncas_climate/ajh/data_analysis_tools/VM.

Dependencies#

cf-plot has the following dependencies:

cf-python, 'cf-python >= 3.17.0'
Cartopy, 'cartopy >= 0.22.0'

With package/environment managers such as conda and pip you can install all dependencies along with the package, but you can also find information on installing these dependencies separately on the links in the list above to the relevant documentation pages of each dependency library.

Sample datasets#

The examples make use of a small number of sample datasets. These are available to view and download from a dedicated directory in the codebase repository. Alternatively, they may be downloaded together as a pair of zip files (split into two parts to fit within GitHub file size limits, total size ~125 MB) available from here (part 1 of 2) and here (part 2 of 2) which can be unzipped together to the full dataset directory as follows:

Accessing all of the datasets for the cf-plot examples#

# Download the split zip parts: either use 'wget' as below or
# download both files from the links above
wget https://github.com/NCAS-CMS/cf-plot/tree/main/docs/source/_downloads/example-datasets.z01
wget https://github.com/NCAS-CMS/cf-plot/tree/main/docs/source/_downloads/example-datasets.zip

# Recombine into a single zip (zip -s0 merges split parts)
zip -s0 example-datasets.zip --out all-example-datasets.zip

# Unzip the merged file
unzip all-example-datasets.zip