scikit-gstat 0.2.8

Geostatistical expansion in the scipy style

Info: scikit-gstat needs Python >= 3.5!

How to cite

In case you use SciKit-GStat in other software or scientific publications, please reference this module. It is published and has a DOI. It can be cited as:

Mirko Mälicke, & Helge David Schneider. (2019, November 25). Scikit-GStat 0.2.7: A scipy flavored geostatistical analysis toolbox written in Python. (Version v.0.2.7). Zenodo. http://doi.org/10.5281/zenodo.3552235

Full Documentation

The full documentation can be found at: https://mmaelicke.github.io/scikit-gstat

Description

SciKit-Gstat is a scipy-styled analysis module for geostatistics. It includes two base classes Variogram and OrdinaryKriging. Additionally, various variogram classes inheriting from Variogram are available for solving directional or space-time related tasks. The module makes use of a rich selection of semi-variance estimators and variogram model functions, while being extensible at the same time. The estimators include:

• matheron

• cressie

• dowd

• genton

• entropy

• two experimental ones: quantiles, minmax

The models include:

• sperical

• exponential

• gaussian

• cubic

• stable

• matérn

with all of them in a nugget and no-nugget variation. All the estimator are implemented using numba’s jit decorator. The usage of numba might be subject to change in future versions.

Installation

PyPI:

pip install scikit-gstat

GIT:

git clone https://github.com/mmaelicke/scikit-gstat.git
cd scikit-gstat
pip install -r requirements.txt
pip install -e .

Note: It can happen that the installation of shapely, numba or numpy is failing using pip. Especially on Windows systems. Usually, a missing Dll (see eg. #31) or visual c++ redistributable is the reason. These errors are not caused by pip, scikit-gstat or the respective packages and there are a lot of issues in the shapely and numpy repo concerning these problems. Usually, the best workaround is to install especially shapely independent from scikit-gstat. As far as I know, these problems do not apply if anaconda is used like:

conda install shapely numpy

Usage

The Variogram class needs at least a list of coordiantes and values. All other attributes are set by default. You can easily set up an example by generating some random data:

import numpy as np
import skgstat as skg

coordinates = np.random.gamma(0.7, 2, (30,2))
values = np.random.gamma(2, 2, 30)

V = skg.Variogram(coordinates=coordinates, values=values)
print(V)

spherical Variogram
-------------------
Estimator:    matheron
Range:        1.64
Sill:         5.35
Nugget:       0.00

New Version 0.2

Scikit-gstat was rewritten in major parts. Most of the changes are internal, but the attributes and behaviour of the Variogram has also changed substantially. A detailed description of of the new versions usage will follow. The last version of the old Variogram class, 0.1.8, is kept in the version-0.1.8 branch on GitHub, but not developed any further. Those two versions are not compatible.