Preprocessor

Preprocessor Interface

Constructor

Preprocessor()

Implements an ML preprocessing operation.

Fit/Transform APIs

fit(dataset)	Fit this Preprocessor to the Dataset.
fit_transform(dataset)	Fit this Preprocessor to the Dataset and then transform the Dataset.
transform(dataset)	Transform the given dataset.
transform_batch(data)	Transform a single batch of data.
transform_stats()	Return Dataset stats for the most recent transform call, if any.

Generic Preprocessors

BatchMapper(fn, batch_format[, batch_size])	Apply an arbitrary operation to a dataset.
Chain(*preprocessors)	Combine multiple preprocessors into a single Preprocessor.
Concatenator([output_column_name, include, ...])	Combine numeric columns into a column of type TensorDtype.
SimpleImputer(columns[, strategy, fill_value])	Replace missing values with imputed values.

Categorical Encoders

Categorizer(columns[, dtypes])	Convert columns to pd.CategoricalDtype.
LabelEncoder(label_column)	Encode labels as integer targets.
MultiHotEncoder(columns, *[, max_categories])	Multi-hot encode categorical data.
OneHotEncoder(columns, *[, max_categories])	One-hot encode categorical data.
OrdinalEncoder(columns, *[, encode_lists])	Encode values within columns as ordered integer values.

Feature Scalers

MaxAbsScaler(columns)	Scale each column by its absolute max value.
MinMaxScaler(columns)	Scale each column by its range.
Normalizer(columns[, norm])	Scales each sample to have unit norm.
PowerTransformer(columns, power[, method])	Apply a power transform to make your data more normally distributed.
RobustScaler(columns[, quantile_range])	Scale and translate each column using quantiles.
StandardScaler(columns)	Translate and scale each column by its mean and standard deviation, respectively.

K-Bins Discretizers

CustomKBinsDiscretizer(columns, bins, *[, ...])	Bin values into discrete intervals using custom bin edges.
UniformKBinsDiscretizer(columns, bins, *[, ...])	Bin values into discrete intervals (bins) of uniform width.

Image Preprocessors

TorchVisionPreprocessor(columns, transform)

Apply a TorchVision transform to image columns.

Text Encoders

CountVectorizer(columns[, tokenization_fn, ...])	Count the frequency of tokens in a column of strings.
FeatureHasher(columns, num_features)	Apply the hashing trick to a table that describes token frequencies.
HashingVectorizer(columns, num_features[, ...])	Count the frequency of tokens using the hashing trick.
Tokenizer(columns[, tokenization_fn])	Replace each string with a list of tokens.