Main Content


Generate datasets for training deep-learning-based CHOMP optimizer

Since R2024a


    [trainingData,validationData] = generateSamples(dlchomp,options,folderpath) generates training and validation data samples for training the specified dlCHOMP object using the specified data options and writes the data samples to the specified folder path. For more information about how to generate samples, see Train Deep-Learning-Based CHOMP Optimizer for Motion Planning.

    The generateSamples function requires the Deep Learning Toolbox™.

    [___] = generateSamples(___,Verbose=verbose) stops execution when generateSamples cannot generate samples using the specified DLCHOMP data options and error with recommendations for data option values.

    Input Arguments

    collapse all

    Deep-learning-based CHOMP optimizer, specified as an dlCHOMP object.

    Data sample generation options, specified as a dlCHOMPDataOptions object.

    Folder path to store generated data samples, specified as a string scalar or character array.

    Data Types: string | char

    Enable data sampling tips, specified as a logical 1 (true) or logical 0 (false). When verbose is true, generateSamples stops generating samples when the specified data options can not be used to generate the desired number of new samples.

    Output Arguments

    collapse all

    Training samples dataset, returned as a dlCHOMPDatastore object.

    Validation samples dataset, returned as a dlCHOMPDatastore object.

    Version History

    Introduced in R2024a