Skip to content

Tutorials

To show some use cases, we have created a number of tutorials that show applications of how to use DKRegression. Below is an overview of the different configurations that are used for the tutorials.

Name \(dim(X)\) \(dim(Y)\) Kernel Observation Likelihood Example Description
Step Function 1 1 RBF Univariate Gaussian 1D regression problem fitting to a step function with heteroskedastic data.
Temperature Map 2 1 RBF Univariate Gaussian Generate a temperature map over the US from temperature measurements at airports.
Color Preferences 2 1 RBF Bernoulli Fitting preferences for human preferences in color pairings.
Airfoil 2 2 RBF Multivariate Gaussian Fitting a velocity field of a flow around an airfoil.

You can find the tutorials also in the /tutorials directory of the Github repository. The content is the same as the tutorials shown here, but they are provided in the form of a runnable file.

Additional Requirements

Some of the tutorials require additional packages that are not necessary for running DKRegression, but are specific for the applications in the tutorials. For example, the Temperature Map tutorial walks you through generating a temperature map over the US and uses the basemap package for plotting. For your convenience, you can install all the additional dependencies for the tutorials using the following command:

pip install "dkregression[tutorials]"