tonndata

(To be removed) Convert data to standard neural network cell array form

tonndata will be removed in a future release. For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.

For advice on updating your code, see Version History.

Syntax

[y,wasMatrix] = tonndata(x,columnSamples,cellTime)


Description

[y,wasMatrix] = tonndata(x,columnSamples,cellTime) takes these arguments,

x	Matrix or cell array of matrices
columnSamples	True if original samples are oriented as columns, false if rows
cellTime	True if original samples are columns of a cell array, false if they are stored in a matrix

and returns

y	Original data transformed into standard neural network cell array form
wasMatrix	True if original data was a matrix (as opposed to cell array)

If columnSamples is false, then matrix x or matrices in cell array x will be transposed, so row samples will now be stored as column vectors.

If cellTime is false, then matrix samples will be separated into columns of a cell array so time originally represented as vectors in a matrix will now be represented as columns of a cell array.

The returned value wasMatrix can be used by fromnndata to reverse the transformation.

Examples

Here data consisting of six timesteps of 5-element vectors, originally represented as a matrix with six columns, is converted to standard neural network representation and back.

x = rands(5,6)
columnSamples = true; % samples are by columns.
cellTime = false;     % time-steps in matrix, not cell array.
[y,wasMatrix] = tonndata(x,columnSamples,cellTime)
x2 = fromnndata(y,wasMatrix,columnSamples,cellTime)

Version History

Introduced in R2010b

R2026a: tonndata will be removed

tonndata will be removed in a future release.

To train a neural network with time series data, use one of these options:

• Time Series Modeler app — Interactively design and train neural networks for time series data.

• trainnet — Train neural network. Use this function when you need more flexibility in the network architecture or training algorithm.

To learn more about time series data workflows, see these examples and topics:

• Training Neural Networks with Time Series Data

• Time Series Forecasting Using Deep Learning

• Create and Train NARX Network for Time Series Forecasting

• Long Short-Term Memory Neural Networks

• Example Deep Learning Network Architectures

When you use the trainnet function, most datastores and functions output data in the layout that the network expects and you do not need to reshape the data. If your data is in a different layout than what the network expects, then indicate that your data has a different layout by using the InputDataFormats argument of the trainingOptions function, specifying the data as a minibatchqueue object and specifying the MiniBatchFormat property, or by specifying input data as a formatted dlarray object. Specifying data formats is usually easier than preprocessing the input data.

For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.

Version History

Introduced in R2010b

R2026a: tonndata will be removed

tonndata will be removed in a future release.

To learn more about data processing for neural networks, see these examples and topics:

• Preprocess Data for Domain-Specific Deep Learning Applications

• Datastores for Deep Learning

• Choose an App to Label Ground Truth Data

For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.