Table of Contents

Class UnitDataSet

Namespace
TimeSeriesAnalysis.Dynamic
Assembly
TimeSeriesAnalysis.dll

The data for a porition of a process, containg only one output and one or multiple inputs that influence it

public class UnitDataSet
Inheritance
UnitDataSet
Inherited Members

Constructors

UnitDataSet(string)

Constructor for data set without inputs - for "autonomous" processes such as sinusoids, rand walks or other disturbancs.

public UnitDataSet(string name = null)

Parameters

name string

optional internal name of dataset

UnitDataSet((double[], DateTime[]), (double[], DateTime[]), string)

Create a dataset for single-input system from two signals that have separate but overlapping time-series(each given as value-date tuples)

public UnitDataSet((double[], DateTime[]) u, (double[], DateTime[]) y_meas, string name = null)

Parameters

u (double[], DateTime[])

tuple of values and dates describing u

y_meas (double[], DateTime[])

tuple of values and dates describing y

name string

name of dataset

UnitDataSet(UnitDataSet)

Create a copy of an existing data set

public UnitDataSet(UnitDataSet otherDataSet)

Parameters

otherDataSet UnitDataSet

UnitDataSet(UnitDataSet, int, int)

Create a downsampled copy of an existing data set

public UnitDataSet(UnitDataSet originalDataSet, int downsampleFactor, int keyIndex = 0)

Parameters

originalDataSet UnitDataSet
downsampleFactor int

factor by which to downsample the original dataset

keyIndex int

index around which to center the downsampling

Fields

IndicesToIgnore

Indices that are ignored in Y during fitting.

public List<int> IndicesToIgnore

Field Value

List<int>

Properties

BadDataID

Some systems for storing data do not support "NaN", but instead some other magic value is reserved for indicating that a value is bad or missing.

public double BadDataID { get; set; }

Property Value

double

D

Additve output disturbance D (Y = X+ D)

public double[] D { get; set; }

Property Value

double[]

ProcessName

Name

public string ProcessName { get; }

Property Value

string

Times

Timestamps

public DateTime[] Times { get; set; }

Property Value

DateTime[]

U

Input U (given)

public double[,] U { get; set; }

Property Value

double[,]

U_sim

Input U(simulated) - in the case of PID-control

public double[,] U_sim { get; set; }

Property Value

double[,]

Y_meas

Output Y (measured)

public double[] Y_meas { get; set; }

Property Value

double[]

Y_proc

The simulated output of the process, this is a non-measured variable that excludes the disturbance

public double[] Y_proc { get; set; }

Property Value

double[]

Y_setpoint

Setpoint - (if sub-process includes a PID-controller)

public double[] Y_setpoint { get; set; }

Property Value

double[]

Y_sim

Output Y(simulated)

public double[] Y_sim { get; set; }

Property Value

double[]

Methods

Concat(UnitDataSet)

Appends/concatenate another dataset to the end of this one

public bool Concat(UnitDataSet otherDataSet)

Parameters

otherDataSet UnitDataSet

Returns

bool

CreateTimeStamps(double, DateTime?)

If data is already given to object, this method will fill out the timstamps member

public void CreateTimeStamps(double timeBase_s, DateTime? t0 = null)

Parameters

timeBase_s double

the time in seconds bewteen time samples

t0 DateTime?

the date of the first datapoint in the dataset

DetermineIndicesToIgnore(FittingSpecs)

Tags indices to be removed if either of the output is outside the range defined by [Y_min,Y_max], an input is outside [u_min, umax] or if any data matches badDataId

Results are stored in "IndicesToIgnore", not outputted.

public void DetermineIndicesToIgnore(FittingSpecs fittingSpecs)

Parameters

fittingSpecs FittingSpecs

GetAverageU()

Get the average value of each input in the dataset. This is useful when defining model local around a working point.

public double[] GetAverageU()

Returns

double[]

an array of averages, each corrsponding to one column of U. Returns null if it was not possible to calculate averages

GetNumDataPoints()

Returns the number of datapoints in the dataset

public int GetNumDataPoints()

Returns

int

GetOversampledFactor(out int)

Identify the oversampled factor of the dataset. The oversamples need to be evenly spread in the dataset. Subsets of oversamples in an otherwise non-oversampled dataset will not be taken into account here.

public double GetOversampledFactor(out int keyIndex)

Parameters

keyIndex int

Returns

double

the average distance between two unique values in the dataset arrays and a key index where the first new value occurs

GetTimeBase()

Gets the time between samples in seconds, returns zero if times are not set

public double GetTimeBase()

Returns

double

GetTimeSpan()

Get the time spanned by the dataset

public TimeSpan GetTimeSpan()

Returns

TimeSpan

The time spanned by the dataset, or null if times are not set

SetU(double[], double[], double[], double[], double[], double[])

Helper to set column of U to uValues.

public bool SetU(double[] uValues1, double[] uValues2 = null, double[] uValues3 = null, double[] uValues4 = null, double[] uValues5 = null, double[] uValues6 = null)

Parameters

uValues1 double[]

first input array

uValues2 double[]

second input array(can be null)

uValues3 double[]

third input array(can be null)

uValues4 double[]

third input array(can be null)

uValues5 double[]

third input array(can be null)

uValues6 double[]

third input array(can be null)

Returns

bool

true if succesful, otherwise false

SubsetInd(int, int)

Create a copy of the data set that is a subset with given start/end indices

public UnitDataSet SubsetInd(int startInd, int endInd)

Parameters

startInd int
endInd int

Returns

UnitDataSet

SubsetPrc(double, double)

Create a subset of the dataset that is defined in terms of percentages

public UnitDataSet SubsetPrc(double startPrc, double endPrc)

Parameters

startPrc double

A value 0-100%. Should be a value smaller than endPrc

endPrc double

A value 0-100%. Should be a value bigger than startPrc

Returns

UnitDataSet
Edit this page