Gaussian Mixture Random variables Set (tutorial)
The Gaussian Mixture Random Variable Set allows to fit experimental data using kernel smoothing in dimension one or above.
In this tutorial, a distribution with two variables is considered. The samples from the distribution are imported from an ASCII file.
Creation of the object
A new Random Variable Set has to be created. The option Import data from CSV file for Gaussian Mixture Random variables Set is selected.
The input file is selected using browse. The data must be organised in columns separated by a tabulation, space, colon or semicolon. Each column correspond to one dimension. Once the file is added, click on Next, a preview of the data used to create the Gaussian Mixture Random Variable Set is shown.
In this example, the data in the ASCII are comma separated. Choose the comma as a separator to display the values.
The coordinates of the distribution must be given a name. In this example, x denotes the first coordinate and y denotes the second coordinate. Note that the name attributed to the variables must not be already worn by another object in the project.
Click then on on Next, the editor of the Gaussian Mixture Random variables Set appears.
The tab Random Variables shows the Random Variables which have been created, associated with the Gaussian Mixture Random Variable Set. The tab Correlation shows the correlation matrix associated with the different Random Variables. The tab Central points shows the coordinates of the central points used for kernel smoothing.
Plot the data
The probability density function associated with the Random Variables can be displayed by clicking on Show preview. The aspect is the following: