mml

Class SeriesModel.Analysis

java.lang.Object

la.util.Series

mml.SeriesModel.Analysis

Enclosing class:

SeriesModel

public abstract static class SeriesModel.Analysis
extends Series

An Analysis of a Scannable Value (Series). Roughly, for a Series of data produce a Series of Models.

Nested Class Summary

Nested classes/interfaces inherited from class la.util.Series

Series.Discrete, Series.Int, Series.Lines, Series.Range, Series.Separator

Field Summary

Fields

Modifier and Type	Field and Description
protected Series	ds The Series produced by the given data Series.

Fields inherited from class la.util.Series

EMPTY

Constructor Summary

Constructors

Constructor and Description
Analysis(Value.Scannable sv) Construct (prepare for) the Analysis of a given Scannable Value sv.

Method Summary

Modifier and Type	Method and Description
abstract void	advance() Advance to the next Model and the next data element.
Value.Tuple	elt() Return a pair being the Model for the current data element and that data element, that is ⟨eltM(), eltD()⟩.
Value	eltD() Return the current element of the data Series; also see elt().
abstract Model	eltM() Return the Model of the current data element; also see elt() and finalModel().
Model	finalModel() If and when the data Series is exhausted, return the Model of a future, hypothetical data element.
boolean	hasSome() Return ds.hasSome() where ds is the given data Series.
double	nl2Pr() The neg log (base 2) probability of the current data element.
double	nlPr() The negative log probability of the current data element.
int	position() Return ds.position() where ds is in the given data Series.
double	pr() The probability of the current data element.
double	remainingNlPr() The total nlPr() of any remaining elements of the Series, including the current element if there is one.

Methods inherited from class la.util.Series

advance, advanceTo, append, asList, error, exhaust, fromInts, fromVector, hasNone, main, merge, merge, singleton, toString, values, wt

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

ds

protected final Series ds

The Series produced by the given data Series.

Constructor Detail

Analysis

public Analysis(Value.Scannable sv)

Construct (prepare for) the Analysis of a given Scannable Value sv.

Method Detail

hasSome

public boolean hasSome()

Return ds.hasSome() where ds is the given data Series.

Specified by:

hasSome in class Series

position

public int position()

Return ds.position() where ds is in the given data Series. When hasNone(), position() is the length of the data Series.

Specified by:

position in class Series

eltM

public abstract Model eltM()

Return the Model of the current data element; also see elt() and finalModel().

finalModel

public Model finalModel()

If and when the data Series is exhausted, return the Model of a future, hypothetical data element. Also see eltM().

eltD

public Value eltD()

Return the current element of the data Series; also see elt().

elt

public Value.Tuple elt()

Return a pair being the Model for the current data element and that data element, that is ⟨eltM(), eltD()⟩. Throw a RuntimeException if there is no data element. Also see finalModel().

Specified by:

elt in class Series

pr

public double pr()

The probability of the current data element.

nlPr

public double nlPr()

The negative log probability of the current data element. Also see remainingNlPr().

nl2Pr

public double nl2Pr()

The neg log (base 2) probability of the current data element.

remainingNlPr

public double remainingNlPr()

The total nlPr() of any remaining elements of the Series, including the current element if there is one.

advance

public abstract void advance()

Advance to the next Model and the next data element.

Specified by:

advance in class Series