landmark_elastic_registration_warping#
- skfda.preprocessing.registration.landmark_elastic_registration_warping(fd, landmarks, *, location=None, grid_points=None)[source]#
Calculate the transformation used in landmark registration.
Let \(t_{ij}\) the time where the sample \(i\) has the feature \(j\) and \(t^*_j\) the new time for the feature. The warping function will transform the new time in the old time, i.e., \(h_i(t^*_j)=t_{ij}\). The registered samples can be obtained as \(x^*_i(t)=x_i(h_i(t))\).
See [1] for a detailed explanation.
- Parameters:
fd (FData) – Functional data object.
landmarks (ArrayLike) – List containing landmarks for each samples.
location (ArrayLike | None) – Defines where the landmarks will be alligned. By default it will be used as location the mean of the landmarks.
grid_points (Union[ArrayLike, Sequence[ArrayLike]] | None) – Grid of points where the functions are evaluated to obtain a discrete representation of the object.
- Returns:
FDataGrid with the warpings function needed to register the functional data object.
- Raises:
ValueError – If the object to be registered has domain dimension greater than 1 or the list of landmarks or locations does not match with the number of samples.
- Return type:
References
Examples
>>> from skfda.datasets import make_multimodal_landmarks >>> from skfda.datasets import make_multimodal_samples >>> from skfda.preprocessing.registration import ( ... landmark_elastic_registration_warping)
We will create a data with landmarks as example
>>> fd = make_multimodal_samples(n_samples=3, n_modes=2, ... random_state=9) >>> landmarks = make_multimodal_landmarks(n_samples=3, n_modes=2, ... random_state=9) >>> landmarks = landmarks.squeeze()
The function will return the corresponding warping function
>>> warping = landmark_elastic_registration_warping(fd, landmarks) >>> warping FDataGrid(...)
The registered function can be obtained using function composition
>>> fd.compose(warping) FDataGrid(...)
