Converters

All registered Converter components in DashAI.

Name	Description
AdditiveChi2Sampler	Approximate the additive chi-squared kernel using Fourier sampling.
BagOfWordsConverter	Convert raw text documents into a matrix of token occurrence counts.
Binarizer	Threshold each feature to produce binary (0/1) values.
CharacterReplacer	Replace or remove a character or substring in all selected text columns.
ColumnRemover	Remove the columns selected in scope from the dataset.
Embedding	HuggingFace embedding converter.
FastICA	Decompose features into statistically independent components using FastICA.
GenericUnivariateSelect	Select features using a configurable univariate statistical test and mode.
IncrementalPCA	Reduce dimensionality using PCA computed incrementally over mini-batches.
KNNImputer	Fill missing values by averaging over the K nearest complete neighbours.
LabelBinarizer	Binarize a label column into a one-vs-all binary indicator matrix.
LabelEncoder	Encode categorical labels as contiguous integer codes in [0, n_classes - 1].
MaxAbsScaler	Scale each feature by its maximum absolute value to the range [-1, 1].
MinMaxScaler	Scale each feature to a fixed range, by default [0, 1].
MissingIndicator	Add binary indicator columns that flag which values were originally missing.
NanRemover	Remove rows that contain NaN (or null-like) values in the scoped columns.
Normalizer	Normalize each sample (row) independently to unit norm.
Nystroem	Approximate a kernel feature map using the Nystroem method.
OneHotEncoder	Encode categorical columns as binary indicator (one-hot) vectors.
OrdinalEncoder	Encode categorical feature columns as integer ordinal codes.
PCA	Reduce dimensionality using Principal Component Analysis (PCA).
PolynomialFeatures	Generate polynomial and interaction features up to a specified degree.
RBFSampler	Approximate the RBF (Gaussian) kernel feature map via random Fourier features.
RandomUnderSamplerConverter	Balance class distribution by randomly discarding majority-class samples.
SMOTEConverter	Balances class distribution by generating synthetic minority-class samples.
SMOTEENNConverter	Hybrid sampler combining SMOTE oversampling with Edited Nearest Neighbours cleaning.
SelectFdr	Select features by controlling the expected False Discovery Rate (FDR).
SelectFpr	Select features whose p-value is below a False Positive Rate threshold.
SelectFwe	Select features while controlling the Family-Wise Error Rate (FWE).
SelectKBest	Select the K highest-scoring features using a univariate statistical test.
SelectPercentile	Select the top percentile of features by a univariate statistical test.
SimpleImputer	Fill missing values using a simple univariate per-column strategy.
SkewedChi2Sampler	Approximate the skewed chi-squared kernel feature map via random Fourier features.
StandardScaler	Standardize features by removing the mean and scaling to unit variance.
TFIDFConverter	Convert raw text documents into a matrix of TF-IDF weighted features.
TokenizerConverter	Converter that tokenizes text and stores each token ID in a separate column.
TruncatedSVD	Reduce dimensionality using Truncated Singular Value Decomposition (LSA).
VarianceThreshold	Remove features whose variance across the training set is below a threshold.