Main Content

Simulink.SimulationData.Dataset

Access logged simulation data or group simulation input data

Description

A Simulink.SimulationData.Dataset object groups related data. Logged simulation results are often grouped in Dataset objects. You can also use a Dataset object to group simulation input data when you load initial states and when you load external input data using Inport blocks or the Signal Editor block.

A Dataset object created by logging simulation data contains one or more elements. Each element contains data for one logged signal, output, data store, or state. Each element is an object, and the type of the object depends on the data it contains.

When you create a Dataset object that groups simulation input data, each element contains data for a signal, bus, or array of buses. You can add data in any format supported by the loading method you use.

Type of InputData Formats

Scalar, vector, or multidimensional signal

Bus

  • Structure of timeseries, timetable, or matlab.io.datastore.SimulationDatastore objects that matches the hierarchy of the bus

  • Simulink.SimulationData.Signal

Array of buses

  • Array of structures

  • Simulink.SimulationData.Signal

Function-call signal

  • N-by-1 vector

  • Simulink.SimulationData.Signal

Creation

Logging simulation data often creates a Simulink.SimulationData.Dataset object, including when you:

  • Log data using signal logging.

  • Log outputs, states, or final states using the Dataset format.

  • Log data stores.

  • Log data to the workspace or a MAT file using the Record block.

  • Log data in Dataset format using a Scope block.

  • Log data using a Floating Scope or Scope Viewer.

To group external input data for a model in a Dataset object, you can:

  • Create an empty Dataset object and add the input data using the addElement function.

  • Use the createInputDataset to create a Dataset object that contains an element for each root-level Inport block or In Bus Element block in a model. Then, specify the data for each element.

  • Use the Signal Editor to interactively create and edit Dataset objects that contain simulation input data. For details, see Create and Edit Signal Data.

You can also create a Dataset object by converting data that uses another format to use the Dataset format. Having all data in a common format can facilitate postprocessing. For details, see Convert Logged Data to Dataset Format.

Description

example

ds = Simulink.SimulationData.Dataset creates an empty, unnamed Dataset object to which you can add elements. Use this syntax to manually create a Dataset object that contains external input data to load using Inport or In Bus Element blocks.

ds = Simulink.SimulationData.Dataset(dataToConvert) creates an unnamed Dataset object that contains one or more elements that contain the data dataToConvert. Use this syntax to convert data that uses another format to use the Dataset format. For details, see Convert Logged Data to Dataset Format.

The conversion process only converts data for one input at a time. To convert data for multiple variables, convert each variable one at a time, then use the concat function to combine the resulting Dataset objects.

ds = Simulink.SimulationData.Dataset(dataToConvert,"DatasetName",dsName) creates a Dataset object with the name specified by dsName that contains one or more elements that contain the data dataToConvert. Use this syntax to convert data that uses another format to use the Dataset format. For details, see Convert Logged Data to Dataset Format.

Input Arguments

expand all

Data to convert to Dataset format, specified as a timeseries object, a timetable, a structure, an array, or a Simulink.ModelDataLogs object. The results of conversion depend on the format of the input data.

Input DataConversion Results
timeseries object or timetable

Dataset object that contains one Simulink.SimulationData.Signal object with the timeseries or timetable data in its Values property.

ModelDataLogs object

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input object.

Structure that matches the Structure with time logging format.

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input structure.

Structure that matches the Structure logging format

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input structure.

Because the input does not contain time data, the conversion creates a time vector for each signal that starts at 0 and uses a sampling interval of 1.

Array

Dataset object that contains one Simulink.SimulationData.Signal object. The Values property of the Signal object contains a timeseries object, and the Data property of the timeseries object contains the entire array.

Because the input does not contain time data, the conversion creates a time vector for the signal that starts at 0 uses a sampling interval of 1.

For more information, see Convert Logged Data to Dataset Format and Dataset Conversion for Logged Data.

Dataset object name, specified as a string or character vector.

Output Arguments

expand all

Dataset object, returned as a Simulink.SimulationData.Dataset object. The Dataset object is empty when you do not specify input arguments. Create an empty Dataset object when you want to group simulation input data in a single variable.

When you specify input arguments, the Dataset object contains one or more elements that contain the input data. The results of conversion depend on the format of the input data.

Input DataConversion Results
timeseries object

Dataset object that contains one Simulink.SimulationData.Signal object with the timeseries data in its Values property.

ModelDataLogs object

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input object.

Structure that matches the Structure with time logging format

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input structure.

Structure that matches the Structure logging format

Dataset object that contains one or more Simulink.SimulationData.Signal objects that contain the data for each signal in the input structure.

Because the input does not contain time data, the conversion creates a time vector for each signal that starts at 0 and uses a sampling interval of 1.

Array

Dataset object that contains one Simulink.SimulationData.Signal object. The Values property of the Signal object contains a timeseries object, and the Data property of the timeseries object contains the entire array.

Because the input does not contain time data, the conversion creates a time vector for the signal that starts at 0 uses a sampling interval of 1.

For more information, see Convert Logged Data to Dataset Format and Dataset Conversion for Logged Data.

Properties

expand all

Dataset object name, specified as a string or character vector. When you create a Dataset object to group simulation input data, you specify the name for the Dataset object. The name you specify does not need to match the name of the variable that contains the Dataset object.

The Name property of Dataset objects created by logging simulation data is set to match the logging variable name specified in the model configuration parameters.

Object Functions

addElementAdd element to end of a Dataset object
concatConcatenate dataset to another dataset
exportToPreviousReleaseSave a Dataset object to a MAT-file you can open in any release
extractTimetableExtract data from Simulink.SimulationData.Dataset or Simulink.SimulationData.Signal objects into timetables
findGet element or collection of elements from dataset
getGet element or collection of elements from dataset
getElementNamesReturn names of all elements in dataset
numElementsGet number of elements in data set
plotPlot data in the Simulation Data Inspector
setElementChange element stored at specified index

Examples

collapse all

Log the outputs of the vdp model and access the logged data.

Open the model. Configure output logging using the Dataset format and simulate the model using the sim function.

open_system('vdp');
out = sim('vdp','SaveOutput','on','SaveFormat','Dataset');

All logged data is returned in a single variable, out, as a Simulink.SimulationOutput object. The SimulationOutput object contains a Simulink.SimulationData.Dataset object that groups each kind of logged data. This example only logs outputs, so there is one Dataset object, yout.

out
out = 
  Simulink.SimulationOutput:

                   yout: [1x1 Simulink.SimulationData.Dataset] 

     SimulationMetadata: [1x1 Simulink.SimulationMetadata] 
           ErrorMessage: [0x0 char] 

Access the Dataset object that includes the logged output data using a dot and the name of the Dataset object. The Dataset object contains a Simulink.SimulationData.Signal object for each output.

outputs = out.yout
outputs = 
Simulink.SimulationData.Dataset 'yout' with 2 elements

                         Name  BlockPath 
                         ____  _________ 
    1  [1x1 Signal]      x1    vdp/Out1 
    2  [1x1 Signal]      x2    vdp/Out2 

  - Use braces { } to access, modify, or add elements using index.

The Signal object has metadata about the signal, including the path to the block and index of the port that produces the signal. Use the getElement function to access the Signal object that contains the data for signal x1 by name. You can also use curly braces({}) to access elements in a Dataset object by index.

outputX1 = getElement(outputs,'x1')
outputX1 = 
  Simulink.SimulationData.Signal
  Package: Simulink.SimulationData

  Properties:
              Name: 'x1'
    PropagatedName: ''
         BlockPath: [1x1 Simulink.SimulationData.BlockPath]
          PortType: 'inport'
         PortIndex: 1
            Values: [1x1 timeseries]


  Methods, Superclasses

The signal data is stored in the Values property of the Signal object as a timeseries object. The time values are in the Time property of the timeseries object. The signal values are in the Data property.

outputValsX1 = outputX1.Values
  timeseries

  Common Properties:
            Name: 'x1'
            Time: [64x1 double]
        TimeInfo: tsdata.timemetadata
            Data: [64x1 double]
        DataInfo: tsdata.datametadata
outputTimesX1 = outputValsX1.Time
outputTimesX1 = 64×1

         0
    0.0001
    0.0006
    0.0031
    0.0157
    0.0785
    0.2844
    0.5407
    0.8788
    1.2788
      ⋮

outputDataX1 = outputValsX1.Data
outputDataX1 = 64×1

    2.0000
    2.0000
    2.0000
    2.0000
    1.9998
    1.9943
    1.9379
    1.8155
    1.5990
    1.2687
      ⋮

You can also access the time values or data values by combining the steps into a single line of code.

outputDataX1 = getElement(out.yout,'x1').Values.Data
outputDataX1 = 64×1

    2.0000
    2.0000
    2.0000
    2.0000
    1.9998
    1.9943
    1.9379
    1.8155
    1.5990
    1.2687
      ⋮

Create data for three simulation input signals and group them in a Dataset object. A simple model loads the contents of the Dataset object using three root-level Inport blocks. Dashboard Scope blocks in the model display each signal created using the loaded input data.

First, create the signal data to load into the model. Use the expression in this example to create the evenly-spaced time vector for an input signal, especially when modeling discrete input signals. MATLAB® supports several other methods for creating an evenly-spaced vector, but other methods can introduce double-precision rounding errors in the time data, which can lead to unexpected simulation results.

sampleTime = 0.01;
numSteps = 1001;
time = sampleTime*(0:numSteps-1);
time = time';

Create signal data for a sine signal, a cosine signal, and a linear signal.

sineVals = sin(2*pi/3*time);
cosVals = cos(2*pi/3*time);
lineVals = time;

Create a timeseries object to contain the data for each signal. Give each timeseries object a descriptive name so signals are easy to identify once they are grouped in the Dataset object.

sineTS = timeseries(sineVals,time,'Name','Sine Wave');
cosTS = timeseries(cosVals,time,'Name','Cosine Wave');
lineTS = timeseries(lineVals,time,'Name','Line');

Create a Dataset object and use the addElement function to add each timeseries object to the Dataset object.

inputData = Simulink.SimulationData.Dataset;
inputData.Name = 'inputData';
inputData = addElement(inputData,sineTS);
inputData = addElement(inputData,cosTS);
inputData = addElement(inputData,lineTS)
inputData = 
Simulink.SimulationData.Dataset 'inputData' with 3 elements

                             Name         BlockPath 
                             ___________  _________ 
    1  [1x1 timeseries]      Sine Wave    ''       
    2  [1x1 timeseries]      Cosine Wave  ''       
    3  [1x1 timeseries]      Line         ''       

  - Use braces { } to access, modify, or add elements using index.

When you load external input data using root-level Inport blocks, you specify the data to load using the Input parameter in the Model Configuration Parameters on the Data Import/Export pane. Open the model LoadInputDataset and see that the Input parameter is specified as inputData.

open_system('LoadInputDataset.slx');

Simulate the model. The Dashboard Scope block connected to the first Inport block shows the sine signal, the Dashboard Scope block connected to the second Inport block shows the cosine signal, and the Dashboard Scope block connected to the third Inport block shows the linear signal.

out = sim('LoadInputDataset.slx');

You can swap the order of elements in the Dataset object and see the change reflected in how the elements are mapped to the Inport blocks.

inputData{1} = lineTS;
inputData{3} = sineTS
inputData = 
Simulink.SimulationData.Dataset 'inputData' with 3 elements

                             Name         BlockPath 
                             ___________  _________ 
    1  [1x1 timeseries]      Line         ''       
    2  [1x1 timeseries]      Cosine Wave  ''       
    3  [1x1 timeseries]      Sine Wave    ''       

  - Use braces { } to access, modify, or add elements using index.

Simulate the model again. The Dashboard Scope block that displays the first element now shows the line, and the Dashboard Scope block that displays the third element shows the sine wave, reflecting the new order of elements in the Dataset object.

out = sim('LoadInputDataset.slx');

Tips

  • You can use curly braces ({}) to access, add, or modify an element in a Dataset object by index.

  • To access, add, or modify an element of a Dataset object by name, use the getElement, addElement, and setElement functions.

  • When you group simulation inputs using Dataset objects, you can use the Signal Editor or the Signal Editor block to easily change which Dataset object provides input for the simulation.

  • When you load external input data from a Dataset object using root-level Inport blocks, you can use the Root Inport Mapper to map each element in the Dataset object to an Inport block by block name, block path, signal name, or port order. You can also write your own function to map the data. For more information, see Map Root Inport Signal Data.

  • When you save data in a Dataset object to a MAT file, consider saving to a Version 7.3 MAT file if the contents of the Dataset object are too large to fit in memory. You can create a Simulink.SimulationData.DatasetRef that references a Dataset object in a MAT file without loading the data into memory for data processing or for loading big data. For more information, see Load Big Data for Simulations.

  • To save a Dataset object to a MAT file that you can open in an earlier release, use the exportToPreviousRelease function instead of the save function.

  • You can write your own reader to import data from a file into a Dataset object in the base workspace, the model workspace, or the Signal Editor using the Simulink.io.FileType class. For more information, see Create Custom File Type for Import to Signal Editor.

Introduced in R2011a