Class AbstractDataIndexer

Namespace

NOpenNLP.Tools.Ml.Model

Assembly

NOpenNLP.Tools.dll

Abstract class for collecting event and context counts used in training.

public abstract class AbstractDataIndexer : IDataIndexer

Inheritance

object

AbstractDataIndexer

Implements

IDataIndexer

Derived

OnePassDataIndexer

TwoPassDataIndexer

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Fields

CUTOFF_DEFAULT

public const int CUTOFF_DEFAULT = 5

Field Value

int

CUTOFF_PARAM

public const string CUTOFF_PARAM = "Cutoff"

Field Value

string

SORT_DEFAULT

public const bool SORT_DEFAULT = true

Field Value

bool

SORT_PARAM

public const string SORT_PARAM = "sort"

Field Value

string

contexts

The integer contexts associated with each unique event.

protected int[][] contexts

Field Value

int[][]

numTimesEventsSeen

The number of times an event occured in the training data.

protected int[] numTimesEventsSeen

Field Value

int[]

outcomeLabels

The names of the outcomes.

protected string[] outcomeLabels

Field Value

string[]

outcomeList

The integer outcome associated with each unique event.

protected int[] outcomeList

Field Value

int[]

predCounts

The number of times each predicate occured.

protected int[] predCounts

Field Value

int[]

predLabels

The predicate/context names.

protected string[] predLabels

Field Value

string[]

printMessages

protected bool printMessages

Field Value

bool

reportMap

protected IDictionary<string, string>? reportMap

Field Value

IDictionary<string, string>

trainingParameters

protected TrainingParameters trainingParameters

Field Value

TrainingParameters

Properties

Contexts

Gets the array of predicates seen in each event: a 2-D array whose first dimension is the event index and array this refers to contains the contexts for that event.

public virtual int[][] Contexts { get; }

Property Value

int[][]

NumEvents

Gets the number of total events indexed.

public virtual int NumEvents { get; }

Property Value

int

NumTimesEventsSeen

Gets an array indexed by the event index indicating the number of times a particular event was seen.

public virtual int[] NumTimesEventsSeen { get; }

Property Value

int[]

OutcomeLabels

Gets an array of outcome names indexed by outcome index.

public virtual string[] OutcomeLabels { get; }

Property Value

string[]

OutcomeList

Gets an array indicating the outcome index for each event.

public virtual int[] OutcomeList { get; }

Property Value

int[]

PredCounts

Gets an array of the count of each predicate in the events.

public virtual int[] PredCounts { get; }

Property Value

int[]

PredLabels

Gets an array of predicate/context names indexed by context index. These indices are the value of the array returned by Contexts.

public virtual string[] PredLabels { get; }

Property Value

string[]

Values

Gets the values associated with each event context, or null if integer values are to be used.

public virtual float[][]? Values { get; }

Property Value

float[][]

Methods

Display(string)

protected virtual void Display(string s)

Parameters

s string

Index(IObjectStream<Event>)

Performs the data indexing. Make sure the Init(TrainingParameters, IDictionary<string, string>?) method is called first.

public abstract void Index(IObjectStream<Event> eventStream)

Parameters

eventStream IObjectStream<Event>

a stream of events

Exceptions

IOException

if there is an error during reading

Index(IObjectStream<Event?>, IDictionary<string, int>)

protected virtual IList<ComparableEvent?> Index(IObjectStream<Event?> events, IDictionary<string, int> predicateIndex)

Parameters

events IObjectStream<Event>

predicateIndex IDictionary<string, int>

Returns

IList<ComparableEvent>

Init(TrainingParameters, IDictionary<string, string>?)

Sets parameters used during the data indexing.

public virtual void Init(TrainingParameters indexingParameters, IDictionary<string, string>? reportMap)

Parameters

indexingParameters TrainingParameters

reportMap IDictionary<string, string>

the map to which report entries are written.

SortAndMerge(IList<ComparableEvent?>, bool)

Sorts and uniques the array of comparable events and return the number of unique events. This method will alter the eventsToCompare array -- it does an in place sort, followed by an in place edit to remove duplicates.

protected virtual int SortAndMerge(IList<ComparableEvent?> eventsToCompare, bool sort)

Parameters

eventsToCompare IList<ComparableEvent>

a ComparableEvent list

sort bool

whether to sort and merge duplicates

Returns

int

The number of unique events in the specified list.

Exceptions

InsufficientTrainingDataException

if not enough events are provided

ToIndexedStringArray(IDictionary<string, int>)

Utility method for creating a string array from a map whose keys are labels to be stored in the array and whose values are the indices at which the corresponding labels should be inserted.

protected static string[] ToIndexedStringArray(IDictionary<string, int> labelToIndexMap)

Parameters

labelToIndexMap IDictionary<string, int>

Returns

string[]

Update(string[], IDictionary<string, int>)

Updates the set of predicates and counter with the specified event contexts.

protected static void Update(string[] ec, IDictionary<string, int> counter)

Parameters

ec string[]

The contexts/features which occur in a event.

counter IDictionary<string, int>

The predicate counters.