Merge branch 'release/0.6.1'

GAA-UAM · Dec 23, 2021 · cb4e443 · cb4e443
2 parents 3aedb62 + e5adb5b
commit cb4e443
Show file tree

Hide file tree

Showing 94 changed files with 5,401 additions and 2,197 deletions.
diff --git a/.zenodo.json b/.zenodo.json
@@ -0,0 +1,47 @@
+{
+    'creators': [
+        {
+            'affiliation': 'Universidad Autónoma de Madrid',
+            'name': 'Ramos-Carreño, Carlos',
+            'orcid': '0000-0003-2566-7058'
+        },
+        {
+            'affiliation': 'Universidad Autónoma de Madrid',
+            'name': 'Suárez, Alberto',
+            'orcid': '0000-0003-4534-0909'
+        },
+        {
+            'affiliation': 'Universidad Autónoma de Madrid',
+            'name': 'Torrecilla, José Luis',
+            'orcid': '0000-0003-3719-5190'
+        },
+        {
+            'name': 'Carbajo Berrocal, Miguel'
+        },
+        {
+            'name': 'Marcos Manchón, Pablo'
+        },
+        {
+            'name': 'Pérez Manso, Pablo'
+        },
+        {
+            'name': 'Hernando Bernabé, Amanda'
+        },
+        {
+            'name': 'García Fernández, David'
+        },
+        {
+            'name': 'Hong, Yujian'
+        },
+        {
+            'name': 'Rodríguez-Ponga Eyriès, Pedro Martín'
+        },
+        {
+            'name': 'Sánchez Romero, Álvaro'
+        },
+        {
+            'name': 'Petrunina, Elena'
+        }
+    ],
+    'license': 'BSD 3-Clause License',
+}
diff --git a/VERSION b/VERSION
@@ -1 +1 @@
-0.6
+0.6.1
diff --git a/docs/Makefile b/docs/Makefile
@@ -55,13 +55,14 @@ clean:
 	rm -rf modules/autosummary
 	rm -rf modules/exploratory/visualization/autosummary
 	rm -rf modules/exploratory/autosummary
+	rm -rf modules/inference/autosummary
 	rm -rf modules/math/autosummary
-	rm -rf modules/preprocessing/autosummary
-	rm -rf modules/representation/autosummary
 	rm -rf modules/misc/autosummary
 	rm -rf modules/ml/autosummary
 	rm -rf modules/ml/clustering/autosummary
-	rm -rf modules/inference/autosummary
+	rm -rf modules/preprocessing/autosummary
+	rm -rf modules/preprocessing/dim_reduction/autosummary
+	rm -rf modules/representation/autosummary
 	rm -rf backreferences
 
 .PHONY: html

diff --git a/docs/conf.py b/docs/conf.py
@@ -20,6 +20,7 @@
 
 import os
 import sys
+import warnings
 
 import pkg_resources
 # -- Extensions to the  Napoleon GoogleDocstring class ---------------------
@@ -274,6 +275,13 @@ def __call__(self, filename: str) -> str:
     'within_subsection_order': SkfdaExplicitSubOrder,
 }
 
+warnings.filterwarnings(
+    "ignore",
+    category=UserWarning,
+    message='Matplotlib is currently using agg, which is a'
+    ' non-GUI backend, so cannot show the figure.',
+)
+
 autosummary_generate = True
 autodoc_typehints = "description"
 napoleon_use_rtype = True

diff --git a/docs/modules/exploratory/outliers.rst b/docs/modules/exploratory/outliers.rst
@@ -10,8 +10,8 @@ identify the outliers.
 Each of the outlier detection methods in scikit-fda has the same API as the outlier detection methods of
 `scikit-learn <https://scikit-learn.org/stable/modules/outlier_detection.html>`_.
 
-Interquartile Range Outlier Detector
-------------------------------------
+Boxplot Outlier Detector
+------------------------
 
 One of the most common ways of outlier detection is given by the functional data boxplot. An observation is marked
 as an outlier if it has points :math:`1.5 \cdot IQR` times outside the region containing the deepest 50% of the curves
@@ -20,19 +20,20 @@ as an outlier if it has points :math:`1.5 \cdot IQR` times outside the region co
 .. autosummary::
    :toctree: autosummary
 
-   skfda.exploratory.outliers.IQROutlierDetector
+   skfda.exploratory.outliers.BoxplotOutlierDetector
 
 
-DirectionalOutlierDetector
+MSPlotOutlierDetector
 --------------------------
 
-Other more novel way of outlier detection takes into account the magnitude and shape of the curves. Curves which have
+Other more novel way of outlier detection is the one presented in the Magnitude-Shape plot, or
+MS-plot. It takes into account the magnitude and shape of the curves. Curves which have
 a very different shape or magnitude are considered outliers.
 
 .. autosummary::
    :toctree: autosummary
 
-   skfda.exploratory.outliers.DirectionalOutlierDetector
+   skfda.exploratory.outliers.MSPlotOutlierDetector
 
 For this method, it is necessary to compute the mean and variation of the directional outlyingness, which can be done
 with the following function.

diff --git a/docs/modules/exploratory/stats.rst b/docs/modules/exploratory/stats.rst
@@ -18,6 +18,7 @@ measure of the location or central tendency of :term:`functional data`.
    skfda.exploratory.stats.trim_mean
    skfda.exploratory.stats.depth_based_median
    skfda.exploratory.stats.geometric_median
+   skfda.exploratory.stats.fisher_rao_karcher_mean
 
 Dispersion
 ----------

diff --git a/docs/modules/misc.rst b/docs/modules/misc.rst
@@ -3,9 +3,54 @@ Miscellaneous
 
 Miscellaneous functions and objects.
 
+This module groups classes and functions useful to work with functional data
+but which do not belong to other categories.
+
+Mathematical operations
+-----------------------
+
+Some math operations between functional data objects are directly available
+in this module.
+The most important ones are the ones that efficiently compute the inner
+product between functions:
+
+.. autosummary::
+   :toctree: autosummary
+
+   skfda.misc.inner_product
+   skfda.misc.inner_product_matrix
+
+A concept related with the inner product is that of the cosine similarity
+between functions:
+
+.. autosummary::
+   :toctree: autosummary
+
+   skfda.misc.cosine_similarity
+   skfda.misc.cosine_similarity_matrix
+
+Submodules
+----------
+
+In addition the following modules provide useful functionality to work with
+functional data:
+
+- :doc:`misc/covariances`: Contains covariance functions to use with
+  and :func:`~skfda.datasets.make_gaussian_process`
+- :doc:`misc/metrics`: Contains functional data metrics, suitable to being
+  used with several metric-based machine-learning tools.
+- :doc:`misc/operators`: Contains operators, or functions over functions.
+- :doc:`misc/regularization`: Contains regularization functions, usable in
+  contexts such as
+  :class:`linear regression <skfda.ml.regression.LinearRegression>`,
+  :class:`FPCA <skfda.preprocessing.dim_reduction.feature_extraction.FPCA>`,
+  or :class:`basis smoothing <skfda.preprocessing.smoothing.BasisSmoother>`.
+
+
 .. toctree::
    :maxdepth: 4
    :caption: Modules:
+   :hidden:
 
    misc/covariances
    misc/metrics

diff --git a/docs/modules/misc/covariances.rst b/docs/modules/misc/covariances.rst
@@ -14,4 +14,5 @@ processes. These functions can be used as covariances in
    skfda.misc.covariances.Gaussian
    skfda.misc.covariances.Linear
    skfda.misc.covariances.Polynomial
+   skfda.misc.covariances.Matern
    skfda.misc.covariances.WhiteNoise
diff --git a/docs/modules/misc/metrics.rst b/docs/modules/misc/metrics.rst
@@ -34,6 +34,18 @@ value of ``p`` must be explicitly passed in each call.
 
    skfda.misc.metrics.lp_norm
    skfda.misc.metrics.lp_distance
+
+Angular distance
+----------------
+
+The angular distance (using the normalized "angle" between functions given
+by the inner product) is also available, and useful in some contexts.
+
+.. autosummary::
+   :toctree: autosummary
+
+   skfda.misc.metrics.angular_distance
+
 
 Elastic distances
 -----------------
@@ -45,9 +57,8 @@ analysis and registration of functional data.
    :toctree: autosummary
 
     skfda.misc.metrics.fisher_rao_distance
-    skfda.misc.metrics.amplitude_distance
-    skfda.misc.metrics.phase_distance
-    skfda.misc.metrics.warping_distance
+    skfda.misc.metrics.fisher_rao_amplitude_distance
+    skfda.misc.metrics.fisher_rao_phase_distance
 
 
 Metric induced by a norm
@@ -75,3 +86,15 @@ of objets. The following class can compute that efficiently:
    :toctree: autosummary
 
    skfda.misc.metrics.PairwiseMetric
+
+
+Transformation metric
+---------------------
+
+Some metrics, such as those based in derivatives, can be expressed as a
+transformation followed by another metric:
+
+.. autosummary::
+   :toctree: autosummary
+
+   skfda.misc.metrics.TransformationMetric
diff --git a/docs/modules/misc/operators.rst b/docs/modules/misc/operators.rst
@@ -11,4 +11,5 @@ The operators that are linear can also be used in the context of
    :toctree: autosummary
 
    skfda.misc.operators.Identity
-   skfda.misc.operators.LinearDifferentialOperator
+   skfda.misc.operators.LinearDifferentialOperator
+   skfda.misc.operators.SRSF
diff --git a/docs/modules/ml/classification.rst b/docs/modules/ml/classification.rst
@@ -5,7 +5,6 @@ Classification
 
 Module with classes to perform classification of functional data.
 
-
 Nearest Neighbors
 -----------------
 
@@ -21,8 +20,29 @@ it is explained the basic usage of these estimators.
 
    skfda.ml.classification.KNeighborsClassifier
    skfda.ml.classification.RadiusNeighborsClassifier
+
+Nearest Centroid
+----------------
+
+This module contains `nearest centroid
+<https://en.wikipedia.org/wiki/Nearest_centroid_classifier>`_ estimators to
+perform classification.
+
+.. autosummary::
+   :toctree: autosummary
+
    skfda.ml.classification.NearestCentroid
    skfda.ml.classification.DTMClassifier
+
+
+Depth
+-----
+
+This module contains depth based estimators to perform classification.
+
+.. autosummary::
+   :toctree: autosummary
+
    skfda.ml.classification.DDClassifier
    skfda.ml.classification.DDGClassifier
    skfda.ml.classification.MaximumDepthClassifier
diff --git a/docs/modules/ml/clustering.rst b/docs/modules/ml/clustering.rst
@@ -43,7 +43,7 @@ clusters given a metric between their elements, in order to cluster together
 elements that are close from each other. This is repeated until a desired
 number of clusters is obtained. The resulting hierarchy of clusters can be
 represented as a tree, called a dendogram. The following hierarchical
-clusterings are supported: 
+clusterings are supported:
 
 .. autosummary::
    :toctree: autosummary

diff --git a/docs/modules/preprocessing/dim_reduction.rst b/docs/modules/preprocessing/dim_reduction.rst
@@ -21,23 +21,24 @@ following:
    skfda.preprocessing.dim_reduction.variable_selection.MaximaHunting
    skfda.preprocessing.dim_reduction.variable_selection.RecursiveMaximaHunting
    skfda.preprocessing.dim_reduction.variable_selection.RKHSVariableSelection
-
+   skfda.preprocessing.dim_reduction.variable_selection.MinimumRedundancyMaximumRelevance
+
 .. toctree::
    :hidden:
    :maxdepth: 4
    :caption: Modules:
 
    dim_reduction/recursive_maxima_hunting
 
-Projection
-----------
-Another way to reduce the dimension is through projection. For example, in
-functional principal component analysis, we project the data samples
-into a smaller sample of functions that preserve the maximum sample
+Feature extraction
+------------------
+Other dimensionality reduction methods construct new features from
+existing ones. For example, in functional principal component
+analysis, we project the data samples into a smaller sample of
+functions that preserve most of the original
 variance.
 
-.. toctree::
-   :maxdepth: 4
-   :caption: Modules:
+.. autosummary::
+   :toctree: autosummary
 
-   dim_reduction/fpca
+   skfda.preprocessing.dim_reduction.feature_extraction.FPCA
diff --git a/docs/modules/preprocessing/dim_reduction/fpca.rst b/docs/modules/preprocessing/dim_reduction/fpca.rst
@@ -21,4 +21,4 @@ FPCA for functional data in both representations
 .. autosummary::
    :toctree: autosummary
 
-   skfda.preprocessing.dim_reduction.projection.FPCA
+   skfda.preprocessing.dim_reduction.feature_extraction.FPCA
diff --git a/docs/modules/preprocessing/registration.rst b/docs/modules/preprocessing/registration.rst
@@ -20,7 +20,7 @@ is shown the basic usage of this method.
 .. autosummary::
    :toctree: autosummary
 
-   skfda.preprocessing.registration.ShiftRegistration
+   skfda.preprocessing.registration.LeastSquaresShiftRegistration
 
 
 Landmark Registration
@@ -36,7 +36,7 @@ by performing a translation in the time scale. See the
 .. autosummary::
    :toctree: autosummary
 
-   skfda.preprocessing.registration.landmark_shift
+   skfda.preprocessing.registration.landmark_shift_registration
    skfda.preprocessing.registration.landmark_shift_deltas
 
 
@@ -47,8 +47,8 @@ See the :ref:`sphx_glr_auto_examples_plot_landmark_registration.py` example.
 .. autosummary::
    :toctree: autosummary
 
-   skfda.preprocessing.registration.landmark_registration
-   skfda.preprocessing.registration.landmark_registration_warping
+   skfda.preprocessing.registration.landmark_elastic_registration
+   skfda.preprocessing.registration.landmark_elastic_registration_warping
 
 
 Elastic Registration
@@ -64,19 +64,7 @@ introduction to this topic along the usage of the corresponding functions.
 .. autosummary::
    :toctree: autosummary
 
-   skfda.preprocessing.registration.ElasticRegistration
-
-
-The module contains some routines related with the elastic registration, making
-a transformation of the sampling, computing different means or distances based
-on the elastic framework.
-
-.. autosummary::
-   :toctree: autosummary
-
-   skfda.preprocessing.registration.elastic.elastic_mean
-   skfda.preprocessing.registration.elastic.warping_mean
-   skfda.preprocessing.registration.elastic.SRSF
+   skfda.preprocessing.registration.FisherRaoElasticRegistration
 
 
 Validation