Installation¶

From a checkout¶

The package is not on conda-forge yet. Until it is, install from a checkout of the repository:

git clone https://github.com/ismip/ismip7-scalar-processing.git
cd ismip7-scalar-processing
conda env create -f ismip7_scalars_env.yml
conda activate ismip7-scalars
python -m pip install --no-deps --no-build-isolation .
ismip7-scalars --version

The last line should print a version. Once a conda-forge package exists, this page will say conda create instead.

What it installs¶

Three commands:

ismip7-scalars processes one experiment (Running the processing)
ismip7-scalars-ensemble processes every experiment in a submission tree (Processing a whole ensemble)
ismip7-scalars-set-params writes a model’s params.nc (File and directory conventions)

and a Python package, ismip7_scalars, whose slc subpackage holds the sea-level methods in The sea-level methods. You can call those from your own analysis:

from ismip7_scalars.slc import slc_vaf

python -m ismip7_scalars does the same as ismip7-scalars, which helps when several environments are on the path at once.

Dependencies¶

ismip7_scalars_env.yml installs these:

Package	Versions	Reason for the bounds
python	3.11 to 3.14	uses tomllib and modern type syntax
numpy	2.1 up to 3	what recent conda-forge netCDF4 builds are built against
netCDF4	1.7 up to 2	every file is read and written through it
isschecker	0.2 up to 1	ships the ISMIP7 data request; see Where the conventions come from

The test suite runs at both ends of every range. If you report a problem, please include the output of conda list for your environment.

MATLAB¶

The MATLAB implementation is not installed by any of this. It is the matlab/scalars.m script in the repository, run from a checkout. See The MATLAB implementation.