sits (version 1.5.2)

sits_stats: Obtain statistics for all sample bands

Description

Most machine learning algorithms require data to be normalized. This applies to the "SVM" method and to all deep learning ones. To normalize the predictors, it is necessary to extract the statistics of each band of the samples. This function computes the 2 of the distribution of each band of the samples. This values are used as minimum and maximum values in the normalization operation performed by the sits_pred_normalize() function.

Usage

sits_stats(samples)

Value

A list with the 2

training data.

Arguments

samples

Time series samples uses as training data.

Author

Gilberto Camara,

Examples

if (sits_run_examples()) {
    stats <- sits_stats(samples_modis_ndvi)
}