data_models.encodings.api

data_models.encodings.api

Classes

Name	Description
CategoricalEncoding	Base class for all categorical encodings.
OneHotEncoding	One-hot encode a category into one binary column per category.
OrdinalEncoding	Encode a category as its integer index `0 .. len(categories) - 1`.
DescriptorEncoding	Encode categories via a per-category descriptor row.

CategoricalEncoding

data_models.encodings.api.CategoricalEncoding()

Base class for all categorical encodings.

Methods

Name	Description
decode	Back-transform encoded columns to categories.
encode	Encode a series of categories into model-input columns.
get_bounds	Lower/upper bounds of the encoded columns for `feature`.
get_names	Final (encoded) column names produced for `feature`.
validate_for_feature	Validate this encoder instance is applicable to `feature`.

decode

data_models.encodings.api.CategoricalEncoding.decode(feature, values)

Back-transform encoded columns to categories.

encode

data_models.encodings.api.CategoricalEncoding.encode(feature, values)

Encode a series of categories into model-input columns.

The returned frame’s columns equal get_names(feature).

get_bounds

data_models.encodings.api.CategoricalEncoding.get_bounds(feature, values=None)

Lower/upper bounds of the encoded columns for feature.

If values is None the optimization bounds (over allowed categories) are returned, else the full bounds over all categories.

get_names

data_models.encodings.api.CategoricalEncoding.get_names(feature)

Final (encoded) column names produced for feature.

validate_for_feature

data_models.encodings.api.CategoricalEncoding.validate_for_feature(feature)

Validate this encoder instance is applicable to feature.

The type-level check (valid_transform_types) only ensures the encoder class is allowed; this instance-level hook rejects encoders whose concrete configuration cannot produce columns for the feature (e.g. a DescriptorEncoding listing static columns the feature does not carry).

OneHotEncoding

data_models.encodings.api.OneHotEncoding()

One-hot encode a category into one binary column per category.

Attributes

Name	Type	Description
drop_first	bool	If True, drop the first category’s column (dummy encoding), yielding `len(categories) - 1` columns.

OrdinalEncoding

data_models.encodings.api.OrdinalEncoding()

Encode a category as its integer index 0 .. len(categories) - 1.

DescriptorEncoding

data_models.encodings.api.DescriptorEncoding()

Encode categories via a per-category descriptor row.

Inherits columns / generators / filter_descriptors / correlation_cutoff from :class:DescriptorSpec. With the default (no static columns listed, no generators) it uses all of the feature’s numeric descriptor columns.

Methods

Name	Description
table	Per-category descriptor table: one row per category, in `categories` order.

table

data_models.encodings.api.DescriptorEncoding.table(feature)

Per-category descriptor table: one row per category, in categories order.

# data_models.encodings.api { #bofire.data_models.encodings.api } `data_models.encodings.api` ## Classes | Name | Description | | --- | --- | | [CategoricalEncoding](#bofire.data_models.encodings.api.CategoricalEncoding) | Base class for all categorical encodings. | | [OneHotEncoding](#bofire.data_models.encodings.api.OneHotEncoding) | One-hot encode a category into one binary column per category. | | [OrdinalEncoding](#bofire.data_models.encodings.api.OrdinalEncoding) | Encode a category as its integer index ``0 .. len(categories) - 1``. | | [DescriptorEncoding](#bofire.data_models.encodings.api.DescriptorEncoding) | Encode categories via a per-category descriptor row. | ### CategoricalEncoding { #bofire.data_models.encodings.api.CategoricalEncoding } ```python data_models.encodings.api.CategoricalEncoding() ``` Base class for all categorical encodings. #### Methods | Name | Description | | --- | --- | | [decode](#bofire.data_models.encodings.api.CategoricalEncoding.decode) | Back-transform encoded columns to categories. | | [encode](#bofire.data_models.encodings.api.CategoricalEncoding.encode) | Encode a series of categories into model-input columns. | | [get_bounds](#bofire.data_models.encodings.api.CategoricalEncoding.get_bounds) | Lower/upper bounds of the encoded columns for ``feature``. | | [get_names](#bofire.data_models.encodings.api.CategoricalEncoding.get_names) | Final (encoded) column names produced for ``feature``. | | [validate_for_feature](#bofire.data_models.encodings.api.CategoricalEncoding.validate_for_feature) | Validate this encoder instance is applicable to ``feature``. | ##### decode { #bofire.data_models.encodings.api.CategoricalEncoding.decode } ```python data_models.encodings.api.CategoricalEncoding.decode(feature, values) ``` Back-transform encoded columns to categories. ##### encode { #bofire.data_models.encodings.api.CategoricalEncoding.encode } ```python data_models.encodings.api.CategoricalEncoding.encode(feature, values) ``` Encode a series of categories into model-input columns. The returned frame's columns equal ``get_names(feature)``. ##### get_bounds { #bofire.data_models.encodings.api.CategoricalEncoding.get_bounds } ```python data_models.encodings.api.CategoricalEncoding.get_bounds(feature, values=None) ``` Lower/upper bounds of the encoded columns for ``feature``. If ``values`` is None the optimization bounds (over allowed categories) are returned, else the full bounds over all categories. ##### get_names { #bofire.data_models.encodings.api.CategoricalEncoding.get_names } ```python data_models.encodings.api.CategoricalEncoding.get_names(feature) ``` Final (encoded) column names produced for ``feature``. ##### validate_for_feature { #bofire.data_models.encodings.api.CategoricalEncoding.validate_for_feature } ```python data_models.encodings.api.CategoricalEncoding.validate_for_feature(feature) ``` Validate this encoder instance is applicable to ``feature``. The type-level check (``valid_transform_types``) only ensures the encoder *class* is allowed; this instance-level hook rejects encoders whose concrete configuration cannot produce columns for the feature (e.g. a ``DescriptorEncoding`` listing static columns the feature does not carry). ### OneHotEncoding { #bofire.data_models.encodings.api.OneHotEncoding } ```python data_models.encodings.api.OneHotEncoding() ``` One-hot encode a category into one binary column per category. #### Attributes {.doc-section .doc-section-attributes} | Name | Type | Description | |------------|--------|-------------------------------------------------------------------------------------------------------| | drop_first | bool | If True, drop the first category's column (dummy encoding), yielding ``len(categories) - 1`` columns. | ### OrdinalEncoding { #bofire.data_models.encodings.api.OrdinalEncoding } ```python data_models.encodings.api.OrdinalEncoding() ``` Encode a category as its integer index ``0 .. len(categories) - 1``. ### DescriptorEncoding { #bofire.data_models.encodings.api.DescriptorEncoding } ```python data_models.encodings.api.DescriptorEncoding() ``` Encode categories via a per-category descriptor row. Inherits ``columns`` / ``generators`` / ``filter_descriptors`` / ``correlation_cutoff`` from :class:`DescriptorSpec`. With the default (no static columns listed, no generators) it uses all of the feature's numeric descriptor columns. #### Methods | Name | Description | | --- | --- | | [table](#bofire.data_models.encodings.api.DescriptorEncoding.table) | Per-category descriptor table: one row per category, in ``categories`` order. | ##### table { #bofire.data_models.encodings.api.DescriptorEncoding.table } ```python data_models.encodings.api.DescriptorEncoding.table(feature) ``` Per-category descriptor table: one row per category, in ``categories`` order.