Class PreprocessingPipeline<T, TInput, TOutput>

Namespace

AiDotNet.Preprocessing

Assembly

AiDotNet.dll

Chains multiple data transformers into a sequential pipeline.

public class PreprocessingPipeline<T, TInput, TOutput> : IDataTransformer<T, TInput, TOutput>

Type Parameters

T

The numeric type for calculations.

TInput

The input data type.

TOutput

The output data type.

Inheritance

object

PreprocessingPipeline<T, TInput, TOutput>

Implements

IDataTransformer<T, TInput, TOutput>

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

A preprocessing pipeline applies transformers in sequence, passing the output of each transformer as input to the next. This enables composable preprocessing workflows similar to sklearn's Pipeline.

For Beginners: Think of a pipeline as a series of steps. For example: 1. First, fill in missing values (SimpleImputer) 2. Then, scale the data (StandardScaler) 3. Finally, create polynomial features (PolynomialFeatures)

The pipeline ensures all these steps happen in the right order.

Constructors

PreprocessingPipeline()

Creates a new empty preprocessing pipeline.

public PreprocessingPipeline()

Properties

ColumnIndices

Gets the column indices this transformer operates on (null for pipelines).

public int[]? ColumnIndices { get; }

Property Value

int[]

Count

Gets the number of steps in the pipeline.

public int Count { get; }

Property Value

int

IsFitted

Gets whether this pipeline has been fitted to data.

public bool IsFitted { get; }

Property Value

bool

Steps

Gets the named steps in the pipeline.

public IReadOnlyList<(string Name, IDataTransformer<T, TInput, TInput> Transformer)> Steps { get; }

Property Value

IReadOnlyList<(string Name, IDataTransformer<T, TInput, TInput> Transformer)>

SupportsInverseTransform

Gets whether this pipeline supports inverse transformation.

public bool SupportsInverseTransform { get; }

Property Value

bool

Remarks

A pipeline supports inverse transform if ALL its transformers support it.

Methods

Add(IDataTransformer<T, TInput, TInput>)

Adds a transformer step to the pipeline.

public PreprocessingPipeline<T, TInput, TOutput> Add(IDataTransformer<T, TInput, TInput> transformer)

Parameters

transformer IDataTransformer<T, TInput, TInput>

The transformer to add.

Returns

PreprocessingPipeline<T, TInput, TOutput>

This pipeline for method chaining.

Add(string, IDataTransformer<T, TInput, TInput>)

Adds a named transformer step to the pipeline.

public PreprocessingPipeline<T, TInput, TOutput> Add(string name, IDataTransformer<T, TInput, TInput> transformer)

Parameters

name string

The name for this step.

transformer IDataTransformer<T, TInput, TInput>

The transformer to add.

Returns

PreprocessingPipeline<T, TInput, TOutput>

This pipeline for method chaining.

Clone()

Creates a clone of this pipeline without fitted state.

public PreprocessingPipeline<T, TInput, TOutput> Clone()

Returns

PreprocessingPipeline<T, TInput, TOutput>

A new unfitted pipeline with the same structure.

Fit(TInput)

Fits all transformers in the pipeline to the training data.

public void Fit(TInput data)

Parameters

data TInput

The training data to fit.

FitTransform(TInput)

Fits the pipeline and transforms the data in a single step.

public TOutput FitTransform(TInput data)

Parameters

data TInput

The data to fit and transform.

Returns

TOutput

The transformed data.

GetFeatureNamesOut(string[]?)

Gets the output feature names after all transformations.

public string[] GetFeatureNamesOut(string[]? inputFeatureNames = null)

Parameters

inputFeatureNames string[]

The input feature names (optional).

Returns

string[]

The output feature names.

GetStep(string)

Gets a transformer step by name.

public IDataTransformer<T, TInput, TInput>? GetStep(string name)

Parameters

name string

The step name.

Returns

IDataTransformer<T, TInput, TInput>

The transformer, or null if not found.

InverseTransform(TOutput)

Inverse transforms data through all pipeline steps in reverse order.

public TInput InverseTransform(TOutput data)

Parameters

data TOutput

The transformed data.

Returns

TInput

The original-scale data.

Exceptions

NotSupportedException

Thrown if any step doesn't support inverse.

InvalidOperationException

Thrown if not fitted.

SetFinalTransformer(IDataTransformer<T, TInput, TOutput>)

Sets the final transformer that may change the output type.

public PreprocessingPipeline<T, TInput, TOutput> SetFinalTransformer(IDataTransformer<T, TInput, TOutput> transformer)

Parameters

transformer IDataTransformer<T, TInput, TOutput>

The final transformer.

Returns

PreprocessingPipeline<T, TInput, TOutput>

This pipeline for method chaining.

Transform(TInput)

Transforms data through all pipeline steps.

public TOutput Transform(TInput data)

Parameters

data TInput

The data to transform.

Returns

TOutput

The transformed data.

Exceptions

InvalidOperationException

Thrown if not fitted.