Source code for openTSNE.sklearn
import openTSNE
import numpy as np
[docs]
class TSNE(openTSNE.TSNE):
__doc__ = openTSNE.TSNE.__doc__
[docs]
def fit(self, X, y=None):
"""Fit X into an embedded space.
Parameters
----------
X: np.ndarray
The data matrix to be embedded.
y : ignored
"""
self.fit_transform(X, y)
return self
[docs]
def fit_transform(self, X, y=None):
"""Fit X into an embedded space and return that transformed output.
Parameters
----------
X: np.ndarray
The data matrix to be embedded.
y : ignored
Returns
-------
np.ndarray
Embedding of the training data in low-dimensional space.
"""
embedding = super().fit(X)
self.embedding_ = embedding
return self.embedding_.view(np.ndarray)
[docs]
def transform(self, X, *args, **kwargs):
"""Apply dimensionality reduction to X.
See :meth:`openTSNE.TSNEEmbedding.transform` for additional parameters.
Parameters
----------
X: np.ndarray
The data matrix to be embedded.
Returns
-------
np.ndarray
Embedding of the training data in low-dimensional space.
"""
embedding = self.embedding_.transform(X, *args, **kwargs)
return embedding.view(np.ndarray)