Skip to content

Convert sparse spatial data to a spatially and temporally resolved 2D/3D data

License

Notifications You must be signed in to change notification settings

r-pound/sparse2spatial

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sparse2Spatial (s2s)

Zenodo DOI

Sparse2Spatial contains routines to convert sparse observations into spatially and temporally resolved datasets using machine learning algorithms.

This package uses packages from the existing Python stack (e.g. dask, xarray, pandas, sklearn, XGBoost). Pull Requests are welcome!

Installation

sparse2spatial is currently only installable from source. To do this, you can either clone the source directory and manually install:

$ git clone https://github.com/tsherwen/sparse2spatial.git
$ cd sparse2spatial
$ python setup.py install

or, you can install via pip directly from git:

$ pip install git+https://github.com/tsherwen/sparse2spatial.git

Usage

Example analysis code for using sparse2spatial is available in the scripts folder, along with predictions for sea-surface iodide, CHBr3, and CH2Br2.

A separate package (TreeSurgeon) is available for plotting output the from sklearn RandomForestRegressors models. Scripts are provided in sparse2spatial for making the input .csv files required by TreeSurgeon.

Work using Sparse2Spatial (s2s)

A subfolder in the scripts folder is present per species for work using s2s. This currently includes predictions for sea-surface iodide, CHBr3, and CH2Br2.

Publications using sparse2spatial are detailed below:

  • Research paper on predicting sea-surface iodide using machine learning

For details on this work please see the paper referenced below.

Sherwen, T., Chance, R. J., Tinel, L., Ellis, D., Evans, M. J., and Carpenter, L. J.: A machine learning based global sea-surface iodide distribution, Earth Syst. Sci. Data, https://doi.org/10.5194/essd-2019-40, 1-40, 2019.

A file to process the of the csv file of observational data used by the above paper is also included in the scripts/Iodide folder. The observational data can be found at the archived location below.

Chance R.; Tinel L.; Sherwen T.; Baker A.; Bell T.; Brindle J.; Campos M.L.A.M.; Croot P.; Ducklow H.; He P.; Hoogakker B.; Hopkins F.E.; Hughes C.; Jickells T.; Loades D.; Reyes Macaya D.A.; Mahajan A.S.; Malin G.; Phillips D.P.; Sinha A.K.; Sarkar A.; Roberts I.J.; Roy R.; Song X.; Winklebauer H.A.; Wuttig K.; Yang M.; Zhou P.; Carpenter L.J. (2019). Global sea-surface iodide observations, 1967-2018. British Oceanographic Data Centre - Natural Environment Research Council, UK. https://doi.org/10/czhx

License

Copyright (c) 2018 Tomas Sherwen

This work is licensed under a permissive MIT License.

Contact

Tomas Sherwen - tomas.sherwen@york.ac.uk

About

Convert sparse spatial data to a spatially and temporally resolved 2D/3D data

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 99.9%
  • Roff 0.1%