Class BadDataFinder

For dynamic model identification in system identification, bad data points will create "spurious dynamics" that it is especially important to filter out, otherwise it may destory identification of dynamic terms.

Inheritance
object
BadDataFinder
Inherited Members
Namespace: TimeSeriesAnalysis.Dynamic
Assembly: TimeSeriesAnalysis.dll
Syntax
public static class BadDataFinder

Methods

| Edit this page View Source

GetAllBadIndices(double[], double)

Get all the values which are NaN or equal to the badValueIndicatingValue for a single vector/array

Declaration
public static List<int> GetAllBadIndices(double[] inData, double badValueIndicatingValue)
Parameters
Type Name Description
double[] inData
double badValueIndicatingValue
Returns
Type Description
List<int>
| Edit this page View Source

GetAllBadIndices(TimeSeriesDataSet, double)

Get all the values which are NaN or equal to the badValueIndicatingValue for any and all datapoints in an entire TimeSeriesDataSet

Declaration
public static List<int> GetAllBadIndices(TimeSeriesDataSet inputData, double badValueIndicatingValue)
Parameters
Type Name Description
TimeSeriesDataSet inputData
double badValueIndicatingValue
Returns
Type Description
List<int>
| Edit this page View Source

GetAllBadIndicesPlussNext(double[], double)

Get the bad value indices AND the indices trailing them. This is useful when considering difference equations that require the values both at index k and k-1 to perform identification.

Declaration
public static List<int> GetAllBadIndicesPlussNext(double[] inData, double badValueIndicatingValue)
Parameters
Type Name Description
double[] inData
double badValueIndicatingValue
Returns
Type Description
List<int>