retrieve a raw pointer to the underlying VW::workspace unsafe, only for use in implementation of model merging and its corresponding unit tests

<summary> Parses and creates the example. </summary> <param name="json">The example to parse.</param> <param name="label"> Optional label, taking prec

assume that the passed in object has a field "nativePointer" of type long

std::array::operator[] const is made constexpr in C++14, so this can only be guaranteed to be constexpr when this standard is used.

Internal helper function for UTF-16 to UTF-8 conversion This replaces the deprecated std::wstring_convert and std::codecvt_utf8_utf16

<summary> Returns an instance of TObject from the pool or creates a new instance using the objectFactory if the pool is empty. </summary> <remarks>Thi

Every example that is created with :py:meth:`~vowpalwabbit.Workspace.parse`, :py:meth:`~vowpalwabbit.Workspace.example`, or :py:class:`~vowpalwabbit.E

Helper: learn+finish a csoaa_ldf multi_ex Following established pattern: shared example + action lines, no empty terminator in multi_ex. Empty termina

* Parse a line of text into a VW example * @param line - VW format example string

call clear() on all instantiated types

Construct a NamespaceId Argss: id (NamespaceId/str/integer): id used to create namespace Returns: Namespace

Get specified type of prediction from example .. deprecated:: 9.0.0 Use :py:meth:`~vowpalwabbit.Example.get_prediction` instead Args

Returns a collection of examples for a multiline example learner or a single example for a single example learner. Args:

* The current total sum of the progressive validation loss

<summary> Disposable object factory. </summary> <typeparam name="TSource">The disposable context needed to create objects of <typeparamref name="TObje

C(n,k) = n!/(k!(n-k)!)

Frees the native resources.

* Returns the example to the pool. Must call delete() on example after. * @param example - Parsed example from parse()

Predict with Vowpal Wabbit model Args: X ({array-like, sparse matrix}, shape \(n_samples, n_features or 1\)): Tra