About¶

Purpose and functionality¶

kval is a Python library for working with oceanography data processing and analysis.
It is intended as a rather broad suite of Python tools to work with ocean data and metadata.
- The scope of kval evolves somewhat organically.
One key goal is to ease the post-processing of oceanographic data (focused on shipboard and moored CTD data) from instrument files to publication-ready NetCDF files.

The library works best in a modern Python environment and using Jupyter notebooks.

A warning

kval is in active development - use at your own risk and please be careful!

file: Converting to and from various file format (e.g. read CTD .cnv data to xarray/netCDF)
data: Data post-processing and QC (e.g. CTD post-processing)
metadata: Handling and standardizing metadata according to CF conventions
plots*: Various tools to help make nice (matplotlib) figures
maps: Tools for making maps
geo: Geographical calculations (coordinate transformations, point-to-point distances etc)
ocean: Oceanography-specific tools (e.g. (*) vertical modes, turner angles, wkb scaling, geostrophical calculations)
calc: Various useful functions for numerical calculations.
util: Various backend support functions and wrappers for xarray functionality.
signal: Filtering, spectral analysis, etc.

* Not implemented

Note: These are aspirational guidelines and not always adhered to in the current code structure. We will try to get there!

We will attempt to follow the guidelines from the Scientific Python Library Development Guide.
Releases will be published relatively often, whenever a new functionality has been added. Releases will be archived on zenodo and given a DOI.
[Contributor guidelines TBW]

Written in Python (>=3.10).
Tailored for use in a Jupyter notebook environment.
Data and metadata should be stored in xarray Datasets.
- Intermediate operations using, e.g., numpy or pandas objects are fine, but the end user should interact with Datasets.
Visualizations are produced using matplotlib.
Code should adhere to PEP8 style guide, and all functions should have docstrings.
All functionality should have associated pytest tests.
- Unit tests of individual functions are found in tests/unit_tests/. Its directory structure and contents should mirror that of src/kval.
- Tests of more complex functionality (e.g. processing pipelines using multiple modules) should be put in tests/functional_tests/.
- A collection of sample data to be used in testing is found in tests/test_data/. Should aim to cover a wide range of input data, but we also don’t want this to become too bulky - try to keep file size to a minimum.

All operations that modify data should be recorded in the file metadata.
Wherever possible, and at as early a stage as feasiblee, all available useful metadata should be added to Datasets.
Metadata formatting should adhere to CF and ACDD conventions, supplemented by:
- OceanSITES
- 2021 NPI NetCDF guidelines

The project is maintained by the Oceanography section at the Norwegian Polar Institute.
Development is supported by supported by the project HiAOOS.
External contributions (pull requests, issues, whatever) are very welcome!

Contributions, issues, and PRs are welcome!

About the latest release, 0.4.3:

Various fixes, mostly small bugs.
Removed jupyterlab dependency.
Resolved some version-related issues leading to problems with interactive plots.
Cleaned up the source code a bit.

About the latest release, 0.4.2:

(issue with mpl version - created 0.4.3 to fix) About the latest release, 0.4.1:

About the latest release, 0.4.0:

About the latest release, 0.3.2:

v0.3 used to work out pypi and conda distribution.
v0.3.2 adds a basic documentation infrastructure with minimal documentation.
An updated version with improved documentation and a code overhaul is planned for Spring 2025.

Pull requests, issues, etc are very welcome!