---
myst:
    substitutions:
        macro:
            python: 1-7,14-80,84-92,95-137
            cpp: 5-53,57-65,68-109,112
        bloc1:
            python: 113-115,117
            cpp: 85-87,89
        bloc2:
            python: 125-127
            cpp: 97-98
        dedent:
            python: ""
            cpp: "4"
---
# Macro "**modelerTestKriging.{{extension}}**"

## Objective

The idea is to provide a simple example of a kriging usage, and an how to, to produce plots in one 
dimension to represent the results. This example is the one taken from {{metho}} that uses a very 
simple set of six points as training database:

{{
    "```{literalinclude} " + parent_dir +
        "/roottest/uranie/references/modeler/utf-1D-train.dat" 
    + "\n"
    + ":language: none\n"
    + "```"
}}

The aim will be to get a kriging model that describes this dataset and to check its consistency over 
a certain number of points (here 100 points) which will be the testing database:

{{
    "```{literalinclude} " + parent_dir +
        "/roottest/uranie/references/modeler/utf-1D-test.dat" 
    + "\n"
    + ":language: none\n"
    + "```"
}}

In this example, two different correlation functions are tested and the obtained results are compared 
at the end.

## Macro {{uranie}}

{{
    "```{literalinclude} " + parent_dir +
        "/roottest/uranie/doc/modeler/use_cases/" + language + "/modelerTestKriging." + extension
    + "\n"
    + ":language: " + language + "\n"
    + ":lines: " + macro[language] + "\n"
    + "```"
}}

The first function of this macro (called `PrintSolutions`) is a complex part that will not be 
detailed, used to represent the results.

The macro itself starts by reading the training database and storing it in a dataserver. A 
`TGPBuilder` is created with the chosen correlation function and the hyper-parameters are estimation 
by an optimisation procedure in:

{{
    "```{literalinclude} " + parent_dir +
        "/roottest/uranie/doc/modeler/use_cases/" + language + "/modelerTestKriging." + extension 
    + "\n"
    + ":language: " + language + "\n"
    + ":lines: " + bloc1[language] + "\n"
    + ":dedent: " + dedent[language] + "\n"
    + "```"
}}

The last line shows how to build and retrieve the newly created kriging object.

Finally, this kriging model is tested against the training database, thanks to a `TLauncher2` object, 
as following:

{{
    "```{literalinclude} " + parent_dir +
        "/roottest/uranie/doc/modeler/use_cases/" + language + "/modelerTestKriging." + extension 
    + "\n"
    + ":language: " + language + "\n"
    + ":lines: " + bloc2[language] + "\n"
    + ":dedent: " + dedent[language] + "\n"
    + "```"
}}

## Graph

{{
    "```{figure} " + parent_dir +
        "/roottest/build/uranie/doc/modeler/use_cases/" + language + "/kriging/modelerTestKriging.png\n"
    + ":align: center\n"
    + ":name: use_cases_modelerTestKriging\n"
    + figure_scale + "\n"
    + "\n"
    + "Graph of the macro **\"modelerTestKriging." + extension + "\"**" + "\n"
    + "```"
}}