If you want to discover and learn the main features proposed by the gstlearn package, you are at the right place! First of all, you should read the gstlearn overview, then we strongly suggest you to read courses in the order they appear. Each course assumes that the previous ones are known.

All the courses presented here are systematically available under Python (Jupyter Notebooks) or R (R Markdown) platforms. Use the adequate link in the header of each figure.

For people who are not familiar with R language, we provide a short course to learn the R syntax.

Statistics introduction

  1. Basic statistics: Python / R
  2. Random Variables: Python / R
  3. Expectation, Variance and Covariance: Python / R

Using gstlearn

Getting Started (Python / R)

This script is not actually a course. It simply serves to check if gstlearn library has been installed correctly.

Data Base (Python / R)

Description of the Data Base (numerical). How to create a Db (from different sources), to use shortcuts (assessors), to assign roles to variables (locators).

Graphics (Python / R)

Use the library gtlearn.plot (based on matplotlib) to visualize all the objects of gstlearn library

Variography (Python / R)

Calculate variograms on one or several variables of a Db, for any space dimension. Fit a Model (automatic procedure)

Kriging (Python / R)

Perform estimation of a target variable using Kriging: various options (simple and ordinary). Cross-validation.

Multivariate (Python / R)

Estimation using several variables simultaneously: Cokriging, Kriging with trend, Kriging with External Drift

Factorial Kriging Analysis (Python / R)

Filtering images using Factorial Kriging Analysis. Example on a multivariate acquisition

Simulations (Python / R)

Simulations using Turning Bands method: derive the probability of exceeding a threshold. Extension when using an External Drift

Conditional Expectation (Python / R)

Normal transform (Gaussian anamorphosis). Deriving conditional expectation of exceeding a threshold

PluriGaussian simulations (Python / R)

Categorical Simulations using PluriGaussian model. Automatic model fitting. Connectivity test (using acceptation-rejection technique)

SPDE (Python / R)

Using Stochastic Partial Derivative Equation framework to perform Estimation (extension to the use of External Drift)