uniqed.transformers package

Submodules

uniqed.transformers.transformers module

class uniqed.transformers.transformers.TimeDelayEmbedder(d=3, tau=1)

Bases: sklearn.base.BaseEstimator, sklearn.base.TransformerMixin

fit(x, y=None)
transform(x)
fit_transform(x, y=None)

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Xarray-like of shape (n_samples, n_features)

Input samples.

yarray-like of shape (n_samples,) or (n_samples, n_outputs), default=None

Target values (None for unsupervised transformations).

**fit_paramsdict

Additional fit parameters.

X_newndarray array of shape (n_samples, n_features_new)

Transformed array.

_embedding(x, d, tau)

Time delay embedding

Parameters
  • x (numpy.ndarray) – 1D time series

  • d (int) – Embedding dimension

  • tau (int) – Embedding delay

Returns

Embedded time series with [len(x) - (d - 1) * tau, d] shape

Return type

numpy.ndarray

class uniqed.transformers.transformers.TransformYTrue(d=3, tau=1)

Bases: sklearn.base.BaseEstimator, sklearn.base.TransformerMixin

fit(x, y=None)
transform(x)
fit_transform(x, y=None)

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Xarray-like of shape (n_samples, n_features)

Input samples.

yarray-like of shape (n_samples,) or (n_samples, n_outputs), default=None

Target values (None for unsupervised transformations).

**fit_paramsdict

Additional fit parameters.

X_newndarray array of shape (n_samples, n_features_new)

Transformed array.

_transform_y_true(x)

Transforms y_true to shorter version aligned with a specific embedding (d, tau)

Parameters

x (numpy.ndarray) – array with values

Returns

array truncated symmetrically at the begining and at the end

Return type

numpy.ndarray

_get_faketime_axis(d, tau)

Computes a new shifted time-axis for embedded time-series

Parameters
  • embededd_time_series (numpy.ndarray) – embedded time series (n x d) array, time instances as rows

  • embedding_delay (int) – the embedding delay parameter used in the embedding

Returns

numpy.array with new shifted time-axis

uniqed.transformers.transformers.invertit(score, doit=False)

Inverts score if doit is True

Parameters
  • score (np.ndarray) – score to conditionally invert

  • doit (bool) – invert the score or not (default: False)

Returns

inverted or original score

Return type

np.ndarray

uniqed.transformers.transformers._make_result_df(new_time_axis, outlier_score, y_pred, inv_it, prefix='')

Make result dataFrame for detections

Parameters
  • new_time_axis (np.ndarray) – truncated time axis after embedding

  • outlier_score (np.ndarray) – computed outlier scores

  • y_pred (np.ndarray) – predicted class labels (contains -1s and 1s for the two classes)

  • inv_it (bool) – wheather invert the outlier score or not

  • prefix (str) – some prefix to the columns

Returns

DataFrame with results, in the columns are the score, class_label respectively

Return type

pandas.DataFrame

Module contents