.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/miscellaneous/plot_metadata_routing.py"
.. LINE NUMBERS ARE GIVEN BELOW.
.. only:: html
.. note::
:class: sphx-glr-download-link-note
:ref:`Go to the end <sphx_glr_download_auto_examples_miscellaneous_plot_metadata_routing.py>`
to download the full example code. or to run this example in your browser via JupyterLite or Binder
.. rst-class:: sphx-glr-example-title
.. _sphx_glr_auto_examples_miscellaneous_plot_metadata_routing.py:
================
Metadata Routing
================
.. currentmodule:: sklearn
This document shows how you can use the :ref:`metadata routing mechanism
<metadata_routing>` in scikit-learn to route metadata to the estimators,
scorers, and CV splitters consuming them.
To better understand the following document, we need to introduce two concepts:
routers and consumers. A router is an object which forwards some given data and
metadata to other objects. In most cases, a router is a :term:`meta-estimator`,
i.e. an estimator which takes another estimator as a parameter. A function such
as :func:`sklearn.model_selection.cross_validate` which takes an estimator as a
parameter and forwards data and metadata, is also a router.
A consumer, on the other hand, is an object which accepts and uses some given
metadata. For instance, an estimator taking into account ``sample_weight`` in
its :term:`fit` method is a consumer of ``sample_weight``.
It is possible for an object to be both a router and a consumer. For instance,
a meta-estimator may take into account ``sample_weight`` in certain
calculations, but it may also route it to the underlying estimator.
First a few imports and some random data for the rest of the script.
.. GENERATED FROM PYTHON SOURCE LINES 29-33
.. code-block:: Python
# Authors: The scikit-learn developers
# SPDX-License-Identifier: BSD-3-Clause
.. GENERATED FROM PYTHON SOURCE LINES 34-67
.. code-block:: Python
import warnings
from pprint import pprint
import numpy as np
from sklearn import set_config
from sklearn.base import (
BaseEstimator,
ClassifierMixin,
MetaEstimatorMixin,
RegressorMixin,
TransformerMixin,
clone,
)
from sklearn.linear_model import LinearRegression
from sklearn.utils import metadata_routing
from sklearn.utils.metadata_routing import (
MetadataRouter,
MethodMapping,
get_routing_for_object,
process_routing,
)
from sklearn.utils.validation import check_is_fitted
n_samples, n_features = 100, 4
rng = np.random.RandomState(42)
X = rng.rand(n_samples, n_features)
y = rng.randint(0, 2, size=n_samples)
my_groups = rng.randint(0, 10, size=n_samples)
my_weights = rng.rand(n_samples)
my_other_weights = rng.rand(n_samples)
.. GENERATED FROM PYTHON SOURCE LINES 68-69
Metadata routing is only available if explicitly enabled:
.. GENERATED FROM PYTHON SOURCE LINES 69-72
.. code-block:: Python
set_config(enable_metadata_routing=True)
.. GENERATED FROM PYTHON SOURCE LINES 73-74
This utility function is a dummy to check if a metadata is passed:
.. GENERATED FROM PYTHON SOURCE LINES 74-84
.. code-block:: Python
def check_metadata(obj, **kwargs):
for key, value in kwargs.items():
if value is not None:
print(
f"Received {key} of length = {len(value)} in {obj.__class__.__name__}."
)
else:
print(f"{key} is None in {obj.__class__.__name__}.")
.. GENERATED FROM PYTHON SOURCE LINES 85-86
A utility function to nicely print the routing information of an object:
.. GENERATED FROM PYTHON SOURCE LINES 86-90
.. code-block:: Python
def print_routing(obj):
pprint(obj.get_metadata_routing()._serialize())
.. GENERATED FROM PYTHON SOURCE LINES 91-97
Consuming Estimator
-------------------
Here we demonstrate how an estimator can expose the required API to support
metadata routing as a consumer. Imagine a simple classifier accepting
``sample_weight`` as a metadata on its ``fit`` and ``groups`` in its
``predict`` method:
.. GENERATED FROM PYTHON SOURCE LINES 97-112
.. code-block:: Python
class ExampleClassifier(ClassifierMixin, BaseEstimator):
def fit(self, X, y, sample_weight=None):
check_metadata(self, sample_weight=sample_weight)
# all classifiers need to expose a classes_ attribute once they're fit.
self.classes_ = np.array([0, 1])
return self
def predict(self, X, groups=None):
check_metadata(self, groups=groups)
# return a constant value of 1, not a very smart classifier!
return np.ones(len(X))
.. GENERATED FROM PYTHON SOURCE LINES 113-123
The above estimator now has all it needs to consume metadata. This is
accomplished by some magic done in :class:`~base.BaseEstimator`. There are
now three methods exposed by the above class: ``set_fit_request``,
``set_predict_request``, and ``get_metadata_routing``. There is also a
``set_score_request`` for ``sample_weight`` which is present since
:class:`~base.ClassifierMixin` implements a ``score`` method accepting
``sample_weight``. The same applies to regressors which inherit from
:class:`~base.RegressorMixin`.
By default, no metadata is requested, which we can see as:
.. GENERATED FROM PYTHON SOURCE LINES 123-126
.. code-block:: Python
print_routing(ExampleClassifier())
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'fit': {'sample_weight': None},
'predict': {'groups': None},
'score': {'sample_weight': None}}
.. GENERATED FROM PYTHON SOURCE LINES 127-133
The above output means that ``sample_weight`` and ``groups`` are not
requested by `ExampleClassifier`, and if a router is given those metadata, it
should raise an error, since the user has not explicitly set whether they are
required or not. The same is true for ``sample_weight`` in the ``score``
method, which is inherited from :class:`~base.ClassifierMixin`. In order to
explicitly set request values for those metadata, we can use these methods:
.. GENERATED FROM PYTHON SOURCE LINES 133-142
.. code-block:: Python
est = (
ExampleClassifier()
.set_fit_request(sample_weight=False)
.set_predict_request(groups=True)
.set_score_request(sample_weight=False)
)
print_routing(est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'fit': {'sample_weight': False},
'predict': {'groups': True},
'score': {'sample_weight': False}}
.. GENERATED FROM PYTHON SOURCE LINES 143-149
.. note ::
Please note that as long as the above estimator is not used in a
meta-estimator, the user does not need to set any requests for the
metadata and the set values are ignored, since a consumer does not
validate or route given metadata. A simple usage of the above estimator
would work as expected.
.. GENERATED FROM PYTHON SOURCE LINES 149-154
.. code-block:: Python
est = ExampleClassifier()
est.fit(X, y, sample_weight=my_weights)
est.predict(X[:3, :], groups=my_groups)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
Received groups of length = 100 in ExampleClassifier.
array([1., 1., 1.])
.. GENERATED FROM PYTHON SOURCE LINES 155-160
Routing Meta-Estimator
----------------------
Now, we show how to design a meta-estimator to be a router. As a simplified
example, here is a meta-estimator, which doesn't do much other than routing
the metadata.
.. GENERATED FROM PYTHON SOURCE LINES 160-214
.. code-block:: Python
class MetaClassifier(MetaEstimatorMixin, ClassifierMixin, BaseEstimator):
def __init__(self, estimator):
self.estimator = estimator
def get_metadata_routing(self):
# This method defines the routing for this meta-estimator.
# In order to do so, a `MetadataRouter` instance is created, and the
# routing is added to it. More explanations follow below.
router = MetadataRouter(owner=self.__class__.__name__).add(
estimator=self.estimator,
method_mapping=MethodMapping()
.add(caller="fit", callee="fit")
.add(caller="predict", callee="predict")
.add(caller="score", callee="score"),
)
return router
def fit(self, X, y, **fit_params):
# `get_routing_for_object` returns a copy of the `MetadataRouter`
# constructed by the above `get_metadata_routing` method, that is
# internally called.
request_router = get_routing_for_object(self)
# Meta-estimators are responsible for validating the given metadata.
# `method` refers to the parent's method, i.e. `fit` in this example.
request_router.validate_metadata(params=fit_params, method="fit")
# `MetadataRouter.route_params` maps the given metadata to the metadata
# required by the underlying estimator based on the routing information
# defined by the MetadataRouter. The output of type `Bunch` has a key
# for each consuming object and those hold keys for their consuming
# methods, which then contain key for the metadata which should be
# routed to them.
routed_params = request_router.route_params(params=fit_params, caller="fit")
# A sub-estimator is fitted and its classes are attributed to the
# meta-estimator.
self.estimator_ = clone(self.estimator).fit(X, y, **routed_params.estimator.fit)
self.classes_ = self.estimator_.classes_
return self
def predict(self, X, **predict_params):
check_is_fitted(self)
# As in `fit`, we get a copy of the object's MetadataRouter,
request_router = get_routing_for_object(self)
# then we validate the given metadata,
request_router.validate_metadata(params=predict_params, method="predict")
# and then prepare the input to the underlying `predict` method.
routed_params = request_router.route_params(
params=predict_params, caller="predict"
)
return self.estimator_.predict(X, **routed_params.estimator.predict)
.. GENERATED FROM PYTHON SOURCE LINES 215-232
Let's break down different parts of the above code.
First, the :meth:`~utils.metadata_routing.get_routing_for_object` takes our
meta-estimator (``self``) and returns a
:class:`~utils.metadata_routing.MetadataRouter` or, a
:class:`~utils.metadata_routing.MetadataRequest` if the object is a consumer,
based on the output of the estimator's ``get_metadata_routing`` method.
Then in each method, we use the ``route_params`` method to construct a
dictionary of the form ``{"object_name": {"method_name": {"metadata":
value}}}`` to pass to the underlying estimator's method. The ``object_name``
(``estimator`` in the above ``routed_params.estimator.fit`` example) is the
same as the one added in the ``get_metadata_routing``. ``validate_metadata``
makes sure all given metadata are requested to avoid silent bugs.
Next, we illustrate the different behaviors and notably the type of errors
raised.
.. GENERATED FROM PYTHON SOURCE LINES 232-238
.. code-block:: Python
meta_est = MetaClassifier(
estimator=ExampleClassifier().set_fit_request(sample_weight=True)
)
meta_est.fit(X, y, sample_weight=my_weights)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
.. raw:: html
<div class="output_subarea output_html rendered_html output_result">
<style>#sk-container-id-49 {
/* Definition of color scheme common for light and dark mode */
--sklearn-color-text: #000;
--sklearn-color-text-muted: #666;
--sklearn-color-line: gray;
/* Definition of color scheme for unfitted estimators */
--sklearn-color-unfitted-level-0: #fff5e6;
--sklearn-color-unfitted-level-1: #f6e4d2;
--sklearn-color-unfitted-level-2: #ffe0b3;
--sklearn-color-unfitted-level-3: chocolate;
/* Definition of color scheme for fitted estimators */
--sklearn-color-fitted-level-0: #f0f8ff;
--sklearn-color-fitted-level-1: #d4ebff;
--sklearn-color-fitted-level-2: #b3dbfd;
--sklearn-color-fitted-level-3: cornflowerblue;
/* Specific color for light theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-icon: #696969;
@media (prefers-color-scheme: dark) {
/* Redefinition of color scheme for dark theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-icon: #878787;
}
}
#sk-container-id-49 {
color: var(--sklearn-color-text);
}
#sk-container-id-49 pre {
padding: 0;
}
#sk-container-id-49 input.sk-hidden--visually {
border: 0;
clip: rect(1px 1px 1px 1px);
clip: rect(1px, 1px, 1px, 1px);
height: 1px;
margin: -1px;
overflow: hidden;
padding: 0;
position: absolute;
width: 1px;
}
#sk-container-id-49 div.sk-dashed-wrapped {
border: 1px dashed var(--sklearn-color-line);
margin: 0 0.4em 0.5em 0.4em;
box-sizing: border-box;
padding-bottom: 0.4em;
background-color: var(--sklearn-color-background);
}
#sk-container-id-49 div.sk-container {
/* jupyter's `normalize.less` sets `[hidden] { display: none; }`
but bootstrap.min.css set `[hidden] { display: none !important; }`
so we also need the `!important` here to be able to override the
default hidden behavior on the sphinx rendered scikit-learn.org.
See: https://github.com/scikit-learn/scikit-learn/issues/21755 */
display: inline-block !important;
position: relative;
}
#sk-container-id-49 div.sk-text-repr-fallback {
display: none;
}
div.sk-parallel-item,
div.sk-serial,
div.sk-item {
/* draw centered vertical line to link estimators */
background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));
background-size: 2px 100%;
background-repeat: no-repeat;
background-position: center center;
}
/* Parallel-specific style estimator block */
#sk-container-id-49 div.sk-parallel-item::after {
content: "";
width: 100%;
border-bottom: 2px solid var(--sklearn-color-text-on-default-background);
flex-grow: 1;
}
#sk-container-id-49 div.sk-parallel {
display: flex;
align-items: stretch;
justify-content: center;
background-color: var(--sklearn-color-background);
position: relative;
}
#sk-container-id-49 div.sk-parallel-item {
display: flex;
flex-direction: column;
}
#sk-container-id-49 div.sk-parallel-item:first-child::after {
align-self: flex-end;
width: 50%;
}
#sk-container-id-49 div.sk-parallel-item:last-child::after {
align-self: flex-start;
width: 50%;
}
#sk-container-id-49 div.sk-parallel-item:only-child::after {
width: 0;
}
/* Serial-specific style estimator block */
#sk-container-id-49 div.sk-serial {
display: flex;
flex-direction: column;
align-items: center;
background-color: var(--sklearn-color-background);
padding-right: 1em;
padding-left: 1em;
}
/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is
clickable and can be expanded/collapsed.
- Pipeline and ColumnTransformer use this feature and define the default style
- Estimators will overwrite some part of the style using the `sk-estimator` class
*/
/* Pipeline and ColumnTransformer style (default) */
#sk-container-id-49 div.sk-toggleable {
/* Default theme specific background. It is overwritten whether we have a
specific estimator or a Pipeline/ColumnTransformer */
background-color: var(--sklearn-color-background);
}
/* Toggleable label */
#sk-container-id-49 label.sk-toggleable__label {
cursor: pointer;
display: flex;
width: 100%;
margin-bottom: 0;
padding: 0.5em;
box-sizing: border-box;
text-align: center;
align-items: start;
justify-content: space-between;
gap: 0.5em;
}
#sk-container-id-49 label.sk-toggleable__label .caption {
font-size: 0.6rem;
font-weight: lighter;
color: var(--sklearn-color-text-muted);
}
#sk-container-id-49 label.sk-toggleable__label-arrow:before {
/* Arrow on the left of the label */
content: "▸";
float: left;
margin-right: 0.25em;
color: var(--sklearn-color-icon);
}
#sk-container-id-49 label.sk-toggleable__label-arrow:hover:before {
color: var(--sklearn-color-text);
}
/* Toggleable content - dropdown */
#sk-container-id-49 div.sk-toggleable__content {
max-height: 0;
max-width: 0;
overflow: hidden;
text-align: left;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-49 div.sk-toggleable__content.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-49 div.sk-toggleable__content pre {
margin: 0.2em;
border-radius: 0.25em;
color: var(--sklearn-color-text);
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-49 div.sk-toggleable__content.fitted pre {
/* unfitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-49 input.sk-toggleable__control:checked~div.sk-toggleable__content {
/* Expand drop-down */
max-height: 200px;
max-width: 100%;
overflow: auto;
}
#sk-container-id-49 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {
content: "▾";
}
/* Pipeline/ColumnTransformer-specific style */
#sk-container-id-49 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-49 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator-specific style */
/* Colorize estimator box */
#sk-container-id-49 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-49 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
#sk-container-id-49 div.sk-label label.sk-toggleable__label,
#sk-container-id-49 div.sk-label label {
/* The background is the default theme color */
color: var(--sklearn-color-text-on-default-background);
}
/* On hover, darken the color of the background */
#sk-container-id-49 div.sk-label:hover label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
/* Label box, darken color on hover, fitted */
#sk-container-id-49 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator label */
#sk-container-id-49 div.sk-label label {
font-family: monospace;
font-weight: bold;
display: inline-block;
line-height: 1.2em;
}
#sk-container-id-49 div.sk-label-container {
text-align: center;
}
/* Estimator-specific */
#sk-container-id-49 div.sk-estimator {
font-family: monospace;
border: 1px dotted var(--sklearn-color-border-box);
border-radius: 0.25em;
box-sizing: border-box;
margin-bottom: 0.5em;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-49 div.sk-estimator.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
/* on hover */
#sk-container-id-49 div.sk-estimator:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-49 div.sk-estimator.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
/* Specification for estimator info (e.g. "i" and "?") */
/* Common style for "i" and "?" */
.sk-estimator-doc-link,
a:link.sk-estimator-doc-link,
a:visited.sk-estimator-doc-link {
float: right;
font-size: smaller;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1em;
height: 1em;
width: 1em;
text-decoration: none !important;
margin-left: 0.5em;
text-align: center;
/* unfitted */
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
color: var(--sklearn-color-unfitted-level-1);
}
.sk-estimator-doc-link.fitted,
a:link.sk-estimator-doc-link.fitted,
a:visited.sk-estimator-doc-link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
div.sk-estimator:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover,
div.sk-label-container:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover,
div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
/* Span, style for the box shown on hovering the info icon */
.sk-estimator-doc-link span {
display: none;
z-index: 9999;
position: relative;
font-weight: normal;
right: .2ex;
padding: .5ex;
margin: .5ex;
width: min-content;
min-width: 20ex;
max-width: 50ex;
color: var(--sklearn-color-text);
box-shadow: 2pt 2pt 4pt #999;
/* unfitted */
background: var(--sklearn-color-unfitted-level-0);
border: .5pt solid var(--sklearn-color-unfitted-level-3);
}
.sk-estimator-doc-link.fitted span {
/* fitted */
background: var(--sklearn-color-fitted-level-0);
border: var(--sklearn-color-fitted-level-3);
}
.sk-estimator-doc-link:hover span {
display: block;
}
/* "?"-specific style due to the `<a>` HTML tag */
#sk-container-id-49 a.estimator_doc_link {
float: right;
font-size: 1rem;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1rem;
height: 1rem;
width: 1rem;
text-decoration: none;
/* unfitted */
color: var(--sklearn-color-unfitted-level-1);
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
}
#sk-container-id-49 a.estimator_doc_link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
#sk-container-id-49 a.estimator_doc_link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
#sk-container-id-49 a.estimator_doc_link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
}
</style><div id="sk-container-id-49" class="sk-top-container"><div class="sk-text-repr-fallback"><pre>MetaClassifier(estimator=ExampleClassifier())</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class="sk-container" hidden><div class="sk-item sk-dashed-wrapped"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-210" type="checkbox" ><label for="sk-estimator-id-210" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>MetaClassifier</div></div><div><span class="sk-estimator-doc-link fitted">i<span>Fitted</span></span></div></label><div class="sk-toggleable__content fitted"><pre>MetaClassifier(estimator=ExampleClassifier())</pre></div> </div></div><div class="sk-parallel"><div class="sk-parallel-item"><div class="sk-item"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-211" type="checkbox" ><label for="sk-estimator-id-211" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>estimator: ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div><div class="sk-serial"><div class="sk-item"><div class="sk-estimator fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-212" type="checkbox" ><label for="sk-estimator-id-212" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div></div></div></div></div></div></div></div>
</div>
<br />
<br />
.. GENERATED FROM PYTHON SOURCE LINES 239-243
Note that the above example is calling our utility function
`check_metadata()` via the `ExampleClassifier`. It checks that
``sample_weight`` is correctly passed to it. If it is not, like in the
following example, it would print that ``sample_weight`` is ``None``:
.. GENERATED FROM PYTHON SOURCE LINES 243-246
.. code-block:: Python
meta_est.fit(X, y)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
sample_weight is None in ExampleClassifier.
.. raw:: html
<div class="output_subarea output_html rendered_html output_result">
<style>#sk-container-id-50 {
/* Definition of color scheme common for light and dark mode */
--sklearn-color-text: #000;
--sklearn-color-text-muted: #666;
--sklearn-color-line: gray;
/* Definition of color scheme for unfitted estimators */
--sklearn-color-unfitted-level-0: #fff5e6;
--sklearn-color-unfitted-level-1: #f6e4d2;
--sklearn-color-unfitted-level-2: #ffe0b3;
--sklearn-color-unfitted-level-3: chocolate;
/* Definition of color scheme for fitted estimators */
--sklearn-color-fitted-level-0: #f0f8ff;
--sklearn-color-fitted-level-1: #d4ebff;
--sklearn-color-fitted-level-2: #b3dbfd;
--sklearn-color-fitted-level-3: cornflowerblue;
/* Specific color for light theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-icon: #696969;
@media (prefers-color-scheme: dark) {
/* Redefinition of color scheme for dark theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-icon: #878787;
}
}
#sk-container-id-50 {
color: var(--sklearn-color-text);
}
#sk-container-id-50 pre {
padding: 0;
}
#sk-container-id-50 input.sk-hidden--visually {
border: 0;
clip: rect(1px 1px 1px 1px);
clip: rect(1px, 1px, 1px, 1px);
height: 1px;
margin: -1px;
overflow: hidden;
padding: 0;
position: absolute;
width: 1px;
}
#sk-container-id-50 div.sk-dashed-wrapped {
border: 1px dashed var(--sklearn-color-line);
margin: 0 0.4em 0.5em 0.4em;
box-sizing: border-box;
padding-bottom: 0.4em;
background-color: var(--sklearn-color-background);
}
#sk-container-id-50 div.sk-container {
/* jupyter's `normalize.less` sets `[hidden] { display: none; }`
but bootstrap.min.css set `[hidden] { display: none !important; }`
so we also need the `!important` here to be able to override the
default hidden behavior on the sphinx rendered scikit-learn.org.
See: https://github.com/scikit-learn/scikit-learn/issues/21755 */
display: inline-block !important;
position: relative;
}
#sk-container-id-50 div.sk-text-repr-fallback {
display: none;
}
div.sk-parallel-item,
div.sk-serial,
div.sk-item {
/* draw centered vertical line to link estimators */
background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));
background-size: 2px 100%;
background-repeat: no-repeat;
background-position: center center;
}
/* Parallel-specific style estimator block */
#sk-container-id-50 div.sk-parallel-item::after {
content: "";
width: 100%;
border-bottom: 2px solid var(--sklearn-color-text-on-default-background);
flex-grow: 1;
}
#sk-container-id-50 div.sk-parallel {
display: flex;
align-items: stretch;
justify-content: center;
background-color: var(--sklearn-color-background);
position: relative;
}
#sk-container-id-50 div.sk-parallel-item {
display: flex;
flex-direction: column;
}
#sk-container-id-50 div.sk-parallel-item:first-child::after {
align-self: flex-end;
width: 50%;
}
#sk-container-id-50 div.sk-parallel-item:last-child::after {
align-self: flex-start;
width: 50%;
}
#sk-container-id-50 div.sk-parallel-item:only-child::after {
width: 0;
}
/* Serial-specific style estimator block */
#sk-container-id-50 div.sk-serial {
display: flex;
flex-direction: column;
align-items: center;
background-color: var(--sklearn-color-background);
padding-right: 1em;
padding-left: 1em;
}
/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is
clickable and can be expanded/collapsed.
- Pipeline and ColumnTransformer use this feature and define the default style
- Estimators will overwrite some part of the style using the `sk-estimator` class
*/
/* Pipeline and ColumnTransformer style (default) */
#sk-container-id-50 div.sk-toggleable {
/* Default theme specific background. It is overwritten whether we have a
specific estimator or a Pipeline/ColumnTransformer */
background-color: var(--sklearn-color-background);
}
/* Toggleable label */
#sk-container-id-50 label.sk-toggleable__label {
cursor: pointer;
display: flex;
width: 100%;
margin-bottom: 0;
padding: 0.5em;
box-sizing: border-box;
text-align: center;
align-items: start;
justify-content: space-between;
gap: 0.5em;
}
#sk-container-id-50 label.sk-toggleable__label .caption {
font-size: 0.6rem;
font-weight: lighter;
color: var(--sklearn-color-text-muted);
}
#sk-container-id-50 label.sk-toggleable__label-arrow:before {
/* Arrow on the left of the label */
content: "▸";
float: left;
margin-right: 0.25em;
color: var(--sklearn-color-icon);
}
#sk-container-id-50 label.sk-toggleable__label-arrow:hover:before {
color: var(--sklearn-color-text);
}
/* Toggleable content - dropdown */
#sk-container-id-50 div.sk-toggleable__content {
max-height: 0;
max-width: 0;
overflow: hidden;
text-align: left;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-50 div.sk-toggleable__content.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-50 div.sk-toggleable__content pre {
margin: 0.2em;
border-radius: 0.25em;
color: var(--sklearn-color-text);
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-50 div.sk-toggleable__content.fitted pre {
/* unfitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-50 input.sk-toggleable__control:checked~div.sk-toggleable__content {
/* Expand drop-down */
max-height: 200px;
max-width: 100%;
overflow: auto;
}
#sk-container-id-50 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {
content: "▾";
}
/* Pipeline/ColumnTransformer-specific style */
#sk-container-id-50 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-50 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator-specific style */
/* Colorize estimator box */
#sk-container-id-50 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-50 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
#sk-container-id-50 div.sk-label label.sk-toggleable__label,
#sk-container-id-50 div.sk-label label {
/* The background is the default theme color */
color: var(--sklearn-color-text-on-default-background);
}
/* On hover, darken the color of the background */
#sk-container-id-50 div.sk-label:hover label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
/* Label box, darken color on hover, fitted */
#sk-container-id-50 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator label */
#sk-container-id-50 div.sk-label label {
font-family: monospace;
font-weight: bold;
display: inline-block;
line-height: 1.2em;
}
#sk-container-id-50 div.sk-label-container {
text-align: center;
}
/* Estimator-specific */
#sk-container-id-50 div.sk-estimator {
font-family: monospace;
border: 1px dotted var(--sklearn-color-border-box);
border-radius: 0.25em;
box-sizing: border-box;
margin-bottom: 0.5em;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-50 div.sk-estimator.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
/* on hover */
#sk-container-id-50 div.sk-estimator:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-50 div.sk-estimator.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
/* Specification for estimator info (e.g. "i" and "?") */
/* Common style for "i" and "?" */
.sk-estimator-doc-link,
a:link.sk-estimator-doc-link,
a:visited.sk-estimator-doc-link {
float: right;
font-size: smaller;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1em;
height: 1em;
width: 1em;
text-decoration: none !important;
margin-left: 0.5em;
text-align: center;
/* unfitted */
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
color: var(--sklearn-color-unfitted-level-1);
}
.sk-estimator-doc-link.fitted,
a:link.sk-estimator-doc-link.fitted,
a:visited.sk-estimator-doc-link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
div.sk-estimator:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover,
div.sk-label-container:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover,
div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
/* Span, style for the box shown on hovering the info icon */
.sk-estimator-doc-link span {
display: none;
z-index: 9999;
position: relative;
font-weight: normal;
right: .2ex;
padding: .5ex;
margin: .5ex;
width: min-content;
min-width: 20ex;
max-width: 50ex;
color: var(--sklearn-color-text);
box-shadow: 2pt 2pt 4pt #999;
/* unfitted */
background: var(--sklearn-color-unfitted-level-0);
border: .5pt solid var(--sklearn-color-unfitted-level-3);
}
.sk-estimator-doc-link.fitted span {
/* fitted */
background: var(--sklearn-color-fitted-level-0);
border: var(--sklearn-color-fitted-level-3);
}
.sk-estimator-doc-link:hover span {
display: block;
}
/* "?"-specific style due to the `<a>` HTML tag */
#sk-container-id-50 a.estimator_doc_link {
float: right;
font-size: 1rem;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1rem;
height: 1rem;
width: 1rem;
text-decoration: none;
/* unfitted */
color: var(--sklearn-color-unfitted-level-1);
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
}
#sk-container-id-50 a.estimator_doc_link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
#sk-container-id-50 a.estimator_doc_link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
#sk-container-id-50 a.estimator_doc_link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
}
</style><div id="sk-container-id-50" class="sk-top-container"><div class="sk-text-repr-fallback"><pre>MetaClassifier(estimator=ExampleClassifier())</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class="sk-container" hidden><div class="sk-item sk-dashed-wrapped"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-213" type="checkbox" ><label for="sk-estimator-id-213" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>MetaClassifier</div></div><div><span class="sk-estimator-doc-link fitted">i<span>Fitted</span></span></div></label><div class="sk-toggleable__content fitted"><pre>MetaClassifier(estimator=ExampleClassifier())</pre></div> </div></div><div class="sk-parallel"><div class="sk-parallel-item"><div class="sk-item"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-214" type="checkbox" ><label for="sk-estimator-id-214" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>estimator: ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div><div class="sk-serial"><div class="sk-item"><div class="sk-estimator fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-215" type="checkbox" ><label for="sk-estimator-id-215" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div></div></div></div></div></div></div></div>
</div>
<br />
<br />
.. GENERATED FROM PYTHON SOURCE LINES 247-248
If we pass an unknown metadata, an error is raised:
.. GENERATED FROM PYTHON SOURCE LINES 248-253
.. code-block:: Python
try:
meta_est.fit(X, y, test=my_weights)
except TypeError as e:
print(e)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
MetaClassifier.fit got unexpected argument(s) {'test'}, which are not routed to any object.
.. GENERATED FROM PYTHON SOURCE LINES 254-255
And if we pass a metadata which is not explicitly requested:
.. GENERATED FROM PYTHON SOURCE LINES 255-260
.. code-block:: Python
try:
meta_est.fit(X, y, sample_weight=my_weights).predict(X, groups=my_groups)
except ValueError as e:
print(e)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
[groups] are passed but are not explicitly set as requested or not requested for ExampleClassifier.predict, which is used within MetaClassifier.predict. Call `ExampleClassifier.set_predict_request({metadata}=True/False)` for each metadata you want to request/ignore.
.. GENERATED FROM PYTHON SOURCE LINES 261-262
Also, if we explicitly set it as not requested, but it is provided:
.. GENERATED FROM PYTHON SOURCE LINES 262-272
.. code-block:: Python
meta_est = MetaClassifier(
estimator=ExampleClassifier()
.set_fit_request(sample_weight=True)
.set_predict_request(groups=False)
)
try:
meta_est.fit(X, y, sample_weight=my_weights).predict(X[:3, :], groups=my_groups)
except TypeError as e:
print(e)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
MetaClassifier.predict got unexpected argument(s) {'groups'}, which are not routed to any object.
.. GENERATED FROM PYTHON SOURCE LINES 273-283
Another concept to introduce is **aliased metadata**. This is when an
estimator requests a metadata with a different variable name than the default
variable name. For instance, in a setting where there are two estimators in a
pipeline, one could request ``sample_weight1`` and the other
``sample_weight2``. Note that this doesn't change what the estimator expects,
it only tells the meta-estimator how to map the provided metadata to what is
required. Here's an example, where we pass ``aliased_sample_weight`` to the
meta-estimator, but the meta-estimator understands that
``aliased_sample_weight`` is an alias for ``sample_weight``, and passes it as
``sample_weight`` to the underlying estimator:
.. GENERATED FROM PYTHON SOURCE LINES 283-288
.. code-block:: Python
meta_est = MetaClassifier(
estimator=ExampleClassifier().set_fit_request(sample_weight="aliased_sample_weight")
)
meta_est.fit(X, y, aliased_sample_weight=my_weights)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
.. raw:: html
<div class="output_subarea output_html rendered_html output_result">
<style>#sk-container-id-51 {
/* Definition of color scheme common for light and dark mode */
--sklearn-color-text: #000;
--sklearn-color-text-muted: #666;
--sklearn-color-line: gray;
/* Definition of color scheme for unfitted estimators */
--sklearn-color-unfitted-level-0: #fff5e6;
--sklearn-color-unfitted-level-1: #f6e4d2;
--sklearn-color-unfitted-level-2: #ffe0b3;
--sklearn-color-unfitted-level-3: chocolate;
/* Definition of color scheme for fitted estimators */
--sklearn-color-fitted-level-0: #f0f8ff;
--sklearn-color-fitted-level-1: #d4ebff;
--sklearn-color-fitted-level-2: #b3dbfd;
--sklearn-color-fitted-level-3: cornflowerblue;
/* Specific color for light theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-icon: #696969;
@media (prefers-color-scheme: dark) {
/* Redefinition of color scheme for dark theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-icon: #878787;
}
}
#sk-container-id-51 {
color: var(--sklearn-color-text);
}
#sk-container-id-51 pre {
padding: 0;
}
#sk-container-id-51 input.sk-hidden--visually {
border: 0;
clip: rect(1px 1px 1px 1px);
clip: rect(1px, 1px, 1px, 1px);
height: 1px;
margin: -1px;
overflow: hidden;
padding: 0;
position: absolute;
width: 1px;
}
#sk-container-id-51 div.sk-dashed-wrapped {
border: 1px dashed var(--sklearn-color-line);
margin: 0 0.4em 0.5em 0.4em;
box-sizing: border-box;
padding-bottom: 0.4em;
background-color: var(--sklearn-color-background);
}
#sk-container-id-51 div.sk-container {
/* jupyter's `normalize.less` sets `[hidden] { display: none; }`
but bootstrap.min.css set `[hidden] { display: none !important; }`
so we also need the `!important` here to be able to override the
default hidden behavior on the sphinx rendered scikit-learn.org.
See: https://github.com/scikit-learn/scikit-learn/issues/21755 */
display: inline-block !important;
position: relative;
}
#sk-container-id-51 div.sk-text-repr-fallback {
display: none;
}
div.sk-parallel-item,
div.sk-serial,
div.sk-item {
/* draw centered vertical line to link estimators */
background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));
background-size: 2px 100%;
background-repeat: no-repeat;
background-position: center center;
}
/* Parallel-specific style estimator block */
#sk-container-id-51 div.sk-parallel-item::after {
content: "";
width: 100%;
border-bottom: 2px solid var(--sklearn-color-text-on-default-background);
flex-grow: 1;
}
#sk-container-id-51 div.sk-parallel {
display: flex;
align-items: stretch;
justify-content: center;
background-color: var(--sklearn-color-background);
position: relative;
}
#sk-container-id-51 div.sk-parallel-item {
display: flex;
flex-direction: column;
}
#sk-container-id-51 div.sk-parallel-item:first-child::after {
align-self: flex-end;
width: 50%;
}
#sk-container-id-51 div.sk-parallel-item:last-child::after {
align-self: flex-start;
width: 50%;
}
#sk-container-id-51 div.sk-parallel-item:only-child::after {
width: 0;
}
/* Serial-specific style estimator block */
#sk-container-id-51 div.sk-serial {
display: flex;
flex-direction: column;
align-items: center;
background-color: var(--sklearn-color-background);
padding-right: 1em;
padding-left: 1em;
}
/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is
clickable and can be expanded/collapsed.
- Pipeline and ColumnTransformer use this feature and define the default style
- Estimators will overwrite some part of the style using the `sk-estimator` class
*/
/* Pipeline and ColumnTransformer style (default) */
#sk-container-id-51 div.sk-toggleable {
/* Default theme specific background. It is overwritten whether we have a
specific estimator or a Pipeline/ColumnTransformer */
background-color: var(--sklearn-color-background);
}
/* Toggleable label */
#sk-container-id-51 label.sk-toggleable__label {
cursor: pointer;
display: flex;
width: 100%;
margin-bottom: 0;
padding: 0.5em;
box-sizing: border-box;
text-align: center;
align-items: start;
justify-content: space-between;
gap: 0.5em;
}
#sk-container-id-51 label.sk-toggleable__label .caption {
font-size: 0.6rem;
font-weight: lighter;
color: var(--sklearn-color-text-muted);
}
#sk-container-id-51 label.sk-toggleable__label-arrow:before {
/* Arrow on the left of the label */
content: "▸";
float: left;
margin-right: 0.25em;
color: var(--sklearn-color-icon);
}
#sk-container-id-51 label.sk-toggleable__label-arrow:hover:before {
color: var(--sklearn-color-text);
}
/* Toggleable content - dropdown */
#sk-container-id-51 div.sk-toggleable__content {
max-height: 0;
max-width: 0;
overflow: hidden;
text-align: left;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-51 div.sk-toggleable__content.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-51 div.sk-toggleable__content pre {
margin: 0.2em;
border-radius: 0.25em;
color: var(--sklearn-color-text);
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-51 div.sk-toggleable__content.fitted pre {
/* unfitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-51 input.sk-toggleable__control:checked~div.sk-toggleable__content {
/* Expand drop-down */
max-height: 200px;
max-width: 100%;
overflow: auto;
}
#sk-container-id-51 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {
content: "▾";
}
/* Pipeline/ColumnTransformer-specific style */
#sk-container-id-51 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-51 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator-specific style */
/* Colorize estimator box */
#sk-container-id-51 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-51 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
#sk-container-id-51 div.sk-label label.sk-toggleable__label,
#sk-container-id-51 div.sk-label label {
/* The background is the default theme color */
color: var(--sklearn-color-text-on-default-background);
}
/* On hover, darken the color of the background */
#sk-container-id-51 div.sk-label:hover label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
/* Label box, darken color on hover, fitted */
#sk-container-id-51 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator label */
#sk-container-id-51 div.sk-label label {
font-family: monospace;
font-weight: bold;
display: inline-block;
line-height: 1.2em;
}
#sk-container-id-51 div.sk-label-container {
text-align: center;
}
/* Estimator-specific */
#sk-container-id-51 div.sk-estimator {
font-family: monospace;
border: 1px dotted var(--sklearn-color-border-box);
border-radius: 0.25em;
box-sizing: border-box;
margin-bottom: 0.5em;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-51 div.sk-estimator.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
/* on hover */
#sk-container-id-51 div.sk-estimator:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-51 div.sk-estimator.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
/* Specification for estimator info (e.g. "i" and "?") */
/* Common style for "i" and "?" */
.sk-estimator-doc-link,
a:link.sk-estimator-doc-link,
a:visited.sk-estimator-doc-link {
float: right;
font-size: smaller;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1em;
height: 1em;
width: 1em;
text-decoration: none !important;
margin-left: 0.5em;
text-align: center;
/* unfitted */
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
color: var(--sklearn-color-unfitted-level-1);
}
.sk-estimator-doc-link.fitted,
a:link.sk-estimator-doc-link.fitted,
a:visited.sk-estimator-doc-link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
div.sk-estimator:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover,
div.sk-label-container:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover,
div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
/* Span, style for the box shown on hovering the info icon */
.sk-estimator-doc-link span {
display: none;
z-index: 9999;
position: relative;
font-weight: normal;
right: .2ex;
padding: .5ex;
margin: .5ex;
width: min-content;
min-width: 20ex;
max-width: 50ex;
color: var(--sklearn-color-text);
box-shadow: 2pt 2pt 4pt #999;
/* unfitted */
background: var(--sklearn-color-unfitted-level-0);
border: .5pt solid var(--sklearn-color-unfitted-level-3);
}
.sk-estimator-doc-link.fitted span {
/* fitted */
background: var(--sklearn-color-fitted-level-0);
border: var(--sklearn-color-fitted-level-3);
}
.sk-estimator-doc-link:hover span {
display: block;
}
/* "?"-specific style due to the `<a>` HTML tag */
#sk-container-id-51 a.estimator_doc_link {
float: right;
font-size: 1rem;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1rem;
height: 1rem;
width: 1rem;
text-decoration: none;
/* unfitted */
color: var(--sklearn-color-unfitted-level-1);
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
}
#sk-container-id-51 a.estimator_doc_link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
#sk-container-id-51 a.estimator_doc_link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
#sk-container-id-51 a.estimator_doc_link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
}
</style><div id="sk-container-id-51" class="sk-top-container"><div class="sk-text-repr-fallback"><pre>MetaClassifier(estimator=ExampleClassifier())</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class="sk-container" hidden><div class="sk-item sk-dashed-wrapped"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-216" type="checkbox" ><label for="sk-estimator-id-216" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>MetaClassifier</div></div><div><span class="sk-estimator-doc-link fitted">i<span>Fitted</span></span></div></label><div class="sk-toggleable__content fitted"><pre>MetaClassifier(estimator=ExampleClassifier())</pre></div> </div></div><div class="sk-parallel"><div class="sk-parallel-item"><div class="sk-item"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-217" type="checkbox" ><label for="sk-estimator-id-217" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>estimator: ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div><div class="sk-serial"><div class="sk-item"><div class="sk-estimator fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-218" type="checkbox" ><label for="sk-estimator-id-218" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div></div></div></div></div></div></div></div>
</div>
<br />
<br />
.. GENERATED FROM PYTHON SOURCE LINES 289-291
Passing ``sample_weight`` here will fail since it is requested with an
alias and ``sample_weight`` with that name is not requested:
.. GENERATED FROM PYTHON SOURCE LINES 291-296
.. code-block:: Python
try:
meta_est.fit(X, y, sample_weight=my_weights)
except TypeError as e:
print(e)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
MetaClassifier.fit got unexpected argument(s) {'sample_weight'}, which are not routed to any object.
.. GENERATED FROM PYTHON SOURCE LINES 297-306
This leads us to the ``get_metadata_routing``. The way routing works in
scikit-learn is that consumers request what they need, and routers pass that
along. Additionally, a router exposes what it requires itself so that it can
be used inside another router, e.g. a pipeline inside a grid search object.
The output of the ``get_metadata_routing`` which is a dictionary
representation of a :class:`~utils.metadata_routing.MetadataRouter`, includes
the complete tree of requested metadata by all nested objects and their
corresponding method routings, i.e. which method of a sub-estimator is used
in which method of a meta-estimator:
.. GENERATED FROM PYTHON SOURCE LINES 306-309
.. code-block:: Python
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': 'aliased_sample_weight'},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 310-318
As you can see, the only metadata requested for method ``fit`` is
``"sample_weight"`` with ``"aliased_sample_weight"`` as the alias. The
``~utils.metadata_routing.MetadataRouter`` class enables us to easily create
the routing object which would create the output we need for our
``get_metadata_routing``.
In order to understand how aliases work in meta-estimators, imagine our
meta-estimator inside another one:
.. GENERATED FROM PYTHON SOURCE LINES 318-323
.. code-block:: Python
meta_meta_est = MetaClassifier(estimator=meta_est).fit(
X, y, aliased_sample_weight=my_weights
)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleClassifier.
.. GENERATED FROM PYTHON SOURCE LINES 324-338
In the above example, this is how the ``fit`` method of `meta_meta_est`
will call their sub-estimator's ``fit`` methods::
# user feeds `my_weights` as `aliased_sample_weight` into `meta_meta_est`:
meta_meta_est.fit(X, y, aliased_sample_weight=my_weights):
...
# the first sub-estimator (`meta_est`) expects `aliased_sample_weight`
self.estimator_.fit(X, y, aliased_sample_weight=aliased_sample_weight):
...
# the second sub-estimator (`est`) expects `sample_weight`
self.estimator_.fit(X, y, sample_weight=aliased_sample_weight):
...
.. GENERATED FROM PYTHON SOURCE LINES 340-347
Consuming and routing Meta-Estimator
------------------------------------
For a slightly more complex example, consider a meta-estimator that routes
metadata to an underlying estimator as before, but it also uses some metadata
in its own methods. This meta-estimator is a consumer and a router at the
same time. Implementing one is very similar to what we had before, but with a
few tweaks.
.. GENERATED FROM PYTHON SOURCE LINES 347-402
.. code-block:: Python
class RouterConsumerClassifier(MetaEstimatorMixin, ClassifierMixin, BaseEstimator):
def __init__(self, estimator):
self.estimator = estimator
def get_metadata_routing(self):
router = (
MetadataRouter(owner=self.__class__.__name__)
# defining metadata routing request values for usage in the meta-estimator
.add_self_request(self)
# defining metadata routing request values for usage in the sub-estimator
.add(
estimator=self.estimator,
method_mapping=MethodMapping()
.add(caller="fit", callee="fit")
.add(caller="predict", callee="predict")
.add(caller="score", callee="score"),
)
)
return router
# Since `sample_weight` is used and consumed here, it should be defined as
# an explicit argument in the method's signature. All other metadata which
# are only routed, will be passed as `**fit_params`:
def fit(self, X, y, sample_weight, **fit_params):
if self.estimator is None:
raise ValueError("estimator cannot be None!")
check_metadata(self, sample_weight=sample_weight)
# We add `sample_weight` to the `fit_params` dictionary.
if sample_weight is not None:
fit_params["sample_weight"] = sample_weight
request_router = get_routing_for_object(self)
request_router.validate_metadata(params=fit_params, method="fit")
routed_params = request_router.route_params(params=fit_params, caller="fit")
self.estimator_ = clone(self.estimator).fit(X, y, **routed_params.estimator.fit)
self.classes_ = self.estimator_.classes_
return self
def predict(self, X, **predict_params):
check_is_fitted(self)
# As in `fit`, we get a copy of the object's MetadataRouter,
request_router = get_routing_for_object(self)
# we validate the given metadata,
request_router.validate_metadata(params=predict_params, method="predict")
# and then prepare the input to the underlying ``predict`` method.
routed_params = request_router.route_params(
params=predict_params, caller="predict"
)
return self.estimator_.predict(X, **routed_params.estimator.predict)
.. GENERATED FROM PYTHON SOURCE LINES 403-415
The key parts where the above meta-estimator differs from our previous
meta-estimator is accepting ``sample_weight`` explicitly in ``fit`` and
including it in ``fit_params``. Since ``sample_weight`` is an explicit
argument, we can be sure that ``set_fit_request(sample_weight=...)`` is
present for this method. The meta-estimator is both a consumer, as well as a
router of ``sample_weight``.
In ``get_metadata_routing``, we add ``self`` to the routing using
``add_self_request`` to indicate this estimator is consuming
``sample_weight`` as well as being a router; which also adds a
``$self_request`` key to the routing info as illustrated below. Now let's
look at some examples:
.. GENERATED FROM PYTHON SOURCE LINES 417-418
- No metadata requested
.. GENERATED FROM PYTHON SOURCE LINES 418-422
.. code-block:: Python
meta_est = RouterConsumerClassifier(estimator=ExampleClassifier())
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'$self_request': {'fit': {'sample_weight': None},
'score': {'sample_weight': None}},
'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': None},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 423-424
- ``sample_weight`` requested by sub-estimator
.. GENERATED FROM PYTHON SOURCE LINES 424-429
.. code-block:: Python
meta_est = RouterConsumerClassifier(
estimator=ExampleClassifier().set_fit_request(sample_weight=True)
)
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'$self_request': {'fit': {'sample_weight': None},
'score': {'sample_weight': None}},
'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': True},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 430-431
- ``sample_weight`` requested by meta-estimator
.. GENERATED FROM PYTHON SOURCE LINES 431-436
.. code-block:: Python
meta_est = RouterConsumerClassifier(estimator=ExampleClassifier()).set_fit_request(
sample_weight=True
)
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'$self_request': {'fit': {'sample_weight': True},
'score': {'sample_weight': None}},
'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': None},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 437-441
Note the difference in the requested metadata representations above.
- We can also alias the metadata to pass different values to the fit methods
of the meta- and the sub-estimator:
.. GENERATED FROM PYTHON SOURCE LINES 441-447
.. code-block:: Python
meta_est = RouterConsumerClassifier(
estimator=ExampleClassifier().set_fit_request(sample_weight="clf_sample_weight"),
).set_fit_request(sample_weight="meta_clf_sample_weight")
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'$self_request': {'fit': {'sample_weight': 'meta_clf_sample_weight'},
'score': {'sample_weight': None}},
'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': 'clf_sample_weight'},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 448-451
However, ``fit`` of the meta-estimator only needs the alias for the
sub-estimator and addresses their own sample weight as `sample_weight`, since
it doesn't validate and route its own required metadata:
.. GENERATED FROM PYTHON SOURCE LINES 451-453
.. code-block:: Python
meta_est.fit(X, y, sample_weight=my_weights, clf_sample_weight=my_other_weights)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in RouterConsumerClassifier.
Received sample_weight of length = 100 in ExampleClassifier.
.. raw:: html
<div class="output_subarea output_html rendered_html output_result">
<style>#sk-container-id-52 {
/* Definition of color scheme common for light and dark mode */
--sklearn-color-text: #000;
--sklearn-color-text-muted: #666;
--sklearn-color-line: gray;
/* Definition of color scheme for unfitted estimators */
--sklearn-color-unfitted-level-0: #fff5e6;
--sklearn-color-unfitted-level-1: #f6e4d2;
--sklearn-color-unfitted-level-2: #ffe0b3;
--sklearn-color-unfitted-level-3: chocolate;
/* Definition of color scheme for fitted estimators */
--sklearn-color-fitted-level-0: #f0f8ff;
--sklearn-color-fitted-level-1: #d4ebff;
--sklearn-color-fitted-level-2: #b3dbfd;
--sklearn-color-fitted-level-3: cornflowerblue;
/* Specific color for light theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));
--sklearn-color-icon: #696969;
@media (prefers-color-scheme: dark) {
/* Redefinition of color scheme for dark theme */
--sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));
--sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));
--sklearn-color-icon: #878787;
}
}
#sk-container-id-52 {
color: var(--sklearn-color-text);
}
#sk-container-id-52 pre {
padding: 0;
}
#sk-container-id-52 input.sk-hidden--visually {
border: 0;
clip: rect(1px 1px 1px 1px);
clip: rect(1px, 1px, 1px, 1px);
height: 1px;
margin: -1px;
overflow: hidden;
padding: 0;
position: absolute;
width: 1px;
}
#sk-container-id-52 div.sk-dashed-wrapped {
border: 1px dashed var(--sklearn-color-line);
margin: 0 0.4em 0.5em 0.4em;
box-sizing: border-box;
padding-bottom: 0.4em;
background-color: var(--sklearn-color-background);
}
#sk-container-id-52 div.sk-container {
/* jupyter's `normalize.less` sets `[hidden] { display: none; }`
but bootstrap.min.css set `[hidden] { display: none !important; }`
so we also need the `!important` here to be able to override the
default hidden behavior on the sphinx rendered scikit-learn.org.
See: https://github.com/scikit-learn/scikit-learn/issues/21755 */
display: inline-block !important;
position: relative;
}
#sk-container-id-52 div.sk-text-repr-fallback {
display: none;
}
div.sk-parallel-item,
div.sk-serial,
div.sk-item {
/* draw centered vertical line to link estimators */
background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));
background-size: 2px 100%;
background-repeat: no-repeat;
background-position: center center;
}
/* Parallel-specific style estimator block */
#sk-container-id-52 div.sk-parallel-item::after {
content: "";
width: 100%;
border-bottom: 2px solid var(--sklearn-color-text-on-default-background);
flex-grow: 1;
}
#sk-container-id-52 div.sk-parallel {
display: flex;
align-items: stretch;
justify-content: center;
background-color: var(--sklearn-color-background);
position: relative;
}
#sk-container-id-52 div.sk-parallel-item {
display: flex;
flex-direction: column;
}
#sk-container-id-52 div.sk-parallel-item:first-child::after {
align-self: flex-end;
width: 50%;
}
#sk-container-id-52 div.sk-parallel-item:last-child::after {
align-self: flex-start;
width: 50%;
}
#sk-container-id-52 div.sk-parallel-item:only-child::after {
width: 0;
}
/* Serial-specific style estimator block */
#sk-container-id-52 div.sk-serial {
display: flex;
flex-direction: column;
align-items: center;
background-color: var(--sklearn-color-background);
padding-right: 1em;
padding-left: 1em;
}
/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is
clickable and can be expanded/collapsed.
- Pipeline and ColumnTransformer use this feature and define the default style
- Estimators will overwrite some part of the style using the `sk-estimator` class
*/
/* Pipeline and ColumnTransformer style (default) */
#sk-container-id-52 div.sk-toggleable {
/* Default theme specific background. It is overwritten whether we have a
specific estimator or a Pipeline/ColumnTransformer */
background-color: var(--sklearn-color-background);
}
/* Toggleable label */
#sk-container-id-52 label.sk-toggleable__label {
cursor: pointer;
display: flex;
width: 100%;
margin-bottom: 0;
padding: 0.5em;
box-sizing: border-box;
text-align: center;
align-items: start;
justify-content: space-between;
gap: 0.5em;
}
#sk-container-id-52 label.sk-toggleable__label .caption {
font-size: 0.6rem;
font-weight: lighter;
color: var(--sklearn-color-text-muted);
}
#sk-container-id-52 label.sk-toggleable__label-arrow:before {
/* Arrow on the left of the label */
content: "▸";
float: left;
margin-right: 0.25em;
color: var(--sklearn-color-icon);
}
#sk-container-id-52 label.sk-toggleable__label-arrow:hover:before {
color: var(--sklearn-color-text);
}
/* Toggleable content - dropdown */
#sk-container-id-52 div.sk-toggleable__content {
max-height: 0;
max-width: 0;
overflow: hidden;
text-align: left;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-52 div.sk-toggleable__content.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-52 div.sk-toggleable__content pre {
margin: 0.2em;
border-radius: 0.25em;
color: var(--sklearn-color-text);
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-52 div.sk-toggleable__content.fitted pre {
/* unfitted */
background-color: var(--sklearn-color-fitted-level-0);
}
#sk-container-id-52 input.sk-toggleable__control:checked~div.sk-toggleable__content {
/* Expand drop-down */
max-height: 200px;
max-width: 100%;
overflow: auto;
}
#sk-container-id-52 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {
content: "▾";
}
/* Pipeline/ColumnTransformer-specific style */
#sk-container-id-52 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-52 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator-specific style */
/* Colorize estimator box */
#sk-container-id-52 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-52 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
#sk-container-id-52 div.sk-label label.sk-toggleable__label,
#sk-container-id-52 div.sk-label label {
/* The background is the default theme color */
color: var(--sklearn-color-text-on-default-background);
}
/* On hover, darken the color of the background */
#sk-container-id-52 div.sk-label:hover label.sk-toggleable__label {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-unfitted-level-2);
}
/* Label box, darken color on hover, fitted */
#sk-container-id-52 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {
color: var(--sklearn-color-text);
background-color: var(--sklearn-color-fitted-level-2);
}
/* Estimator label */
#sk-container-id-52 div.sk-label label {
font-family: monospace;
font-weight: bold;
display: inline-block;
line-height: 1.2em;
}
#sk-container-id-52 div.sk-label-container {
text-align: center;
}
/* Estimator-specific */
#sk-container-id-52 div.sk-estimator {
font-family: monospace;
border: 1px dotted var(--sklearn-color-border-box);
border-radius: 0.25em;
box-sizing: border-box;
margin-bottom: 0.5em;
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-0);
}
#sk-container-id-52 div.sk-estimator.fitted {
/* fitted */
background-color: var(--sklearn-color-fitted-level-0);
}
/* on hover */
#sk-container-id-52 div.sk-estimator:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-2);
}
#sk-container-id-52 div.sk-estimator.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-2);
}
/* Specification for estimator info (e.g. "i" and "?") */
/* Common style for "i" and "?" */
.sk-estimator-doc-link,
a:link.sk-estimator-doc-link,
a:visited.sk-estimator-doc-link {
float: right;
font-size: smaller;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1em;
height: 1em;
width: 1em;
text-decoration: none !important;
margin-left: 0.5em;
text-align: center;
/* unfitted */
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
color: var(--sklearn-color-unfitted-level-1);
}
.sk-estimator-doc-link.fitted,
a:link.sk-estimator-doc-link.fitted,
a:visited.sk-estimator-doc-link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
div.sk-estimator:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover,
div.sk-label-container:hover .sk-estimator-doc-link:hover,
.sk-estimator-doc-link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover,
div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,
.sk-estimator-doc-link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
/* Span, style for the box shown on hovering the info icon */
.sk-estimator-doc-link span {
display: none;
z-index: 9999;
position: relative;
font-weight: normal;
right: .2ex;
padding: .5ex;
margin: .5ex;
width: min-content;
min-width: 20ex;
max-width: 50ex;
color: var(--sklearn-color-text);
box-shadow: 2pt 2pt 4pt #999;
/* unfitted */
background: var(--sklearn-color-unfitted-level-0);
border: .5pt solid var(--sklearn-color-unfitted-level-3);
}
.sk-estimator-doc-link.fitted span {
/* fitted */
background: var(--sklearn-color-fitted-level-0);
border: var(--sklearn-color-fitted-level-3);
}
.sk-estimator-doc-link:hover span {
display: block;
}
/* "?"-specific style due to the `<a>` HTML tag */
#sk-container-id-52 a.estimator_doc_link {
float: right;
font-size: 1rem;
line-height: 1em;
font-family: monospace;
background-color: var(--sklearn-color-background);
border-radius: 1rem;
height: 1rem;
width: 1rem;
text-decoration: none;
/* unfitted */
color: var(--sklearn-color-unfitted-level-1);
border: var(--sklearn-color-unfitted-level-1) 1pt solid;
}
#sk-container-id-52 a.estimator_doc_link.fitted {
/* fitted */
border: var(--sklearn-color-fitted-level-1) 1pt solid;
color: var(--sklearn-color-fitted-level-1);
}
/* On hover */
#sk-container-id-52 a.estimator_doc_link:hover {
/* unfitted */
background-color: var(--sklearn-color-unfitted-level-3);
color: var(--sklearn-color-background);
text-decoration: none;
}
#sk-container-id-52 a.estimator_doc_link.fitted:hover {
/* fitted */
background-color: var(--sklearn-color-fitted-level-3);
}
</style><div id="sk-container-id-52" class="sk-top-container"><div class="sk-text-repr-fallback"><pre>RouterConsumerClassifier(estimator=ExampleClassifier())</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class="sk-container" hidden><div class="sk-item sk-dashed-wrapped"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-219" type="checkbox" ><label for="sk-estimator-id-219" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>RouterConsumerClassifier</div></div><div><span class="sk-estimator-doc-link fitted">i<span>Fitted</span></span></div></label><div class="sk-toggleable__content fitted"><pre>RouterConsumerClassifier(estimator=ExampleClassifier())</pre></div> </div></div><div class="sk-parallel"><div class="sk-parallel-item"><div class="sk-item"><div class="sk-label-container"><div class="sk-label fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-220" type="checkbox" ><label for="sk-estimator-id-220" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>estimator: ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div><div class="sk-serial"><div class="sk-item"><div class="sk-estimator fitted sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-221" type="checkbox" ><label for="sk-estimator-id-221" class="sk-toggleable__label fitted sk-toggleable__label-arrow"><div><div>ExampleClassifier</div></div></label><div class="sk-toggleable__content fitted"><pre>ExampleClassifier()</pre></div> </div></div></div></div></div></div></div></div></div>
</div>
<br />
<br />
.. GENERATED FROM PYTHON SOURCE LINES 454-458
- Alias only on the sub-estimator:
This is useful when we don't want the meta-estimator to use the metadata, but
the sub-estimator should.
.. GENERATED FROM PYTHON SOURCE LINES 458-462
.. code-block:: Python
meta_est = RouterConsumerClassifier(
estimator=ExampleClassifier().set_fit_request(sample_weight="aliased_sample_weight")
)
print_routing(meta_est)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
{'$self_request': {'fit': {'sample_weight': None},
'score': {'sample_weight': None}},
'estimator': {'mapping': [{'callee': 'fit', 'caller': 'fit'},
{'callee': 'predict', 'caller': 'predict'},
{'callee': 'score', 'caller': 'score'}],
'router': {'fit': {'sample_weight': 'aliased_sample_weight'},
'predict': {'groups': None},
'score': {'sample_weight': None}}}}
.. GENERATED FROM PYTHON SOURCE LINES 463-466
The meta-estimator cannot use `aliased_sample_weight`, because it expects
it passed as `sample_weight`. This would apply even if
`set_fit_request(sample_weight=True)` was set on it.
.. GENERATED FROM PYTHON SOURCE LINES 468-477
Simple Pipeline
---------------
A slightly more complicated use-case is a meta-estimator resembling a
:class:`~pipeline.Pipeline`. Here is a meta-estimator, which accepts a
transformer and a classifier. When calling its `fit` method, it applies the
transformer's `fit` and `transform` before running the classifier on the
transformed data. Upon `predict`, it applies the transformer's `transform`
before predicting with the classifier's `predict` method on the transformed
new data.
.. GENERATED FROM PYTHON SOURCE LINES 477-534
.. code-block:: Python
class SimplePipeline(ClassifierMixin, BaseEstimator):
def __init__(self, transformer, classifier):
self.transformer = transformer
self.classifier = classifier
def get_metadata_routing(self):
router = (
MetadataRouter(owner=self.__class__.__name__)
# We add the routing for the transformer.
.add(
transformer=self.transformer,
method_mapping=MethodMapping()
# The metadata is routed such that it retraces how
# `SimplePipeline` internally calls the transformer's `fit` and
# `transform` methods in its own methods (`fit` and `predict`).
.add(caller="fit", callee="fit")
.add(caller="fit", callee="transform")
.add(caller="predict", callee="transform"),
)
# We add the routing for the classifier.
.add(
classifier=self.classifier,
method_mapping=MethodMapping()
.add(caller="fit", callee="fit")
.add(caller="predict", callee="predict"),
)
)
return router
def fit(self, X, y, **fit_params):
routed_params = process_routing(self, "fit", **fit_params)
self.transformer_ = clone(self.transformer).fit(
X, y, **routed_params.transformer.fit
)
X_transformed = self.transformer_.transform(
X, **routed_params.transformer.transform
)
self.classifier_ = clone(self.classifier).fit(
X_transformed, y, **routed_params.classifier.fit
)
return self
def predict(self, X, **predict_params):
routed_params = process_routing(self, "predict", **predict_params)
X_transformed = self.transformer_.transform(
X, **routed_params.transformer.transform
)
return self.classifier_.predict(
X_transformed, **routed_params.classifier.predict
)
.. GENERATED FROM PYTHON SOURCE LINES 535-551
Note the usage of :class:`~utils.metadata_routing.MethodMapping` to
declare which methods of the child estimator (callee) are used in which
methods of the meta estimator (caller). As you can see, `SimplePipeline` uses
the transformer's ``transform`` and ``fit`` methods in ``fit``, and its
``transform`` method in ``predict``, and that's what you see implemented in
the routing structure of the pipeline class.
Another difference in the above example with the previous ones is the usage
of :func:`~utils.metadata_routing.process_routing`, which processes the input
parameters, does the required validation, and returns the `routed_params`
which we had created in previous examples. This reduces the boilerplate code
a developer needs to write in each meta-estimator's method. Developers are
strongly recommended to use this function unless there is a good reason
against it.
In order to test the above pipeline, let's add an example transformer.
.. GENERATED FROM PYTHON SOURCE LINES 551-566
.. code-block:: Python
class ExampleTransformer(TransformerMixin, BaseEstimator):
def fit(self, X, y, sample_weight=None):
check_metadata(self, sample_weight=sample_weight)
return self
def transform(self, X, groups=None):
check_metadata(self, groups=groups)
return X
def fit_transform(self, X, y, sample_weight=None, groups=None):
return self.fit(X, y, sample_weight).transform(X, groups)
.. GENERATED FROM PYTHON SOURCE LINES 567-576
Note that in the above example, we have implemented ``fit_transform`` which
calls ``fit`` and ``transform`` with the appropriate metadata. This is only
required if ``transform`` accepts metadata, since the default ``fit_transform``
implementation in :class:`~base.TransformerMixin` doesn't pass metadata to
``transform``.
Now we can test our pipeline, and see if metadata is correctly passed around.
This example uses our `SimplePipeline`, our `ExampleTransformer`, and our
`RouterConsumerClassifier` which uses our `ExampleClassifier`.
.. GENERATED FROM PYTHON SOURCE LINES 576-597
.. code-block:: Python
pipe = SimplePipeline(
transformer=ExampleTransformer()
# we set transformer's fit to receive sample_weight
.set_fit_request(sample_weight=True)
# we set transformer's transform to receive groups
.set_transform_request(groups=True),
classifier=RouterConsumerClassifier(
estimator=ExampleClassifier()
# we want this sub-estimator to receive sample_weight in fit
.set_fit_request(sample_weight=True)
# but not groups in predict
.set_predict_request(groups=False),
)
# and we want the meta-estimator to receive sample_weight as well
.set_fit_request(sample_weight=True),
)
pipe.fit(X, y, sample_weight=my_weights, groups=my_groups).predict(
X[:3], groups=my_groups
)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in ExampleTransformer.
Received groups of length = 100 in ExampleTransformer.
Received sample_weight of length = 100 in RouterConsumerClassifier.
Received sample_weight of length = 100 in ExampleClassifier.
Received groups of length = 100 in ExampleTransformer.
groups is None in ExampleClassifier.
array([1., 1., 1.])
.. GENERATED FROM PYTHON SOURCE LINES 598-605
Deprecation / Default Value Change
----------------------------------
In this section we show how one should handle the case where a router becomes
also a consumer, especially when it consumes the same metadata as its
sub-estimator, or a consumer starts consuming a metadata which it wasn't in
an older release. In this case, a warning should be raised for a while, to
let users know the behavior is changed from previous versions.
.. GENERATED FROM PYTHON SOURCE LINES 605-623
.. code-block:: Python
class MetaRegressor(MetaEstimatorMixin, RegressorMixin, BaseEstimator):
def __init__(self, estimator):
self.estimator = estimator
def fit(self, X, y, **fit_params):
routed_params = process_routing(self, "fit", **fit_params)
self.estimator_ = clone(self.estimator).fit(X, y, **routed_params.estimator.fit)
def get_metadata_routing(self):
router = MetadataRouter(owner=self.__class__.__name__).add(
estimator=self.estimator,
method_mapping=MethodMapping().add(caller="fit", callee="fit"),
)
return router
.. GENERATED FROM PYTHON SOURCE LINES 624-626
As explained above, this is a valid usage if `my_weights` aren't supposed
to be passed as `sample_weight` to `MetaRegressor`:
.. GENERATED FROM PYTHON SOURCE LINES 626-631
.. code-block:: Python
reg = MetaRegressor(estimator=LinearRegression().set_fit_request(sample_weight=True))
reg.fit(X, y, sample_weight=my_weights)
.. GENERATED FROM PYTHON SOURCE LINES 632-634
Now imagine we further develop ``MetaRegressor`` and it now also *consumes*
``sample_weight``:
.. GENERATED FROM PYTHON SOURCE LINES 634-663
.. code-block:: Python
class WeightedMetaRegressor(MetaEstimatorMixin, RegressorMixin, BaseEstimator):
# show warning to remind user to explicitly set the value with
# `.set_{method}_request(sample_weight={boolean})`
__metadata_request__fit = {"sample_weight": metadata_routing.WARN}
def __init__(self, estimator):
self.estimator = estimator
def fit(self, X, y, sample_weight=None, **fit_params):
routed_params = process_routing(
self, "fit", sample_weight=sample_weight, **fit_params
)
check_metadata(self, sample_weight=sample_weight)
self.estimator_ = clone(self.estimator).fit(X, y, **routed_params.estimator.fit)
def get_metadata_routing(self):
router = (
MetadataRouter(owner=self.__class__.__name__)
.add_self_request(self)
.add(
estimator=self.estimator,
method_mapping=MethodMapping().add(caller="fit", callee="fit"),
)
)
return router
.. GENERATED FROM PYTHON SOURCE LINES 664-667
The above implementation is almost the same as ``MetaRegressor``, and
because of the default request value defined in ``__metadata_request__fit``
there is a warning raised when fitted.
.. GENERATED FROM PYTHON SOURCE LINES 667-676
.. code-block:: Python
with warnings.catch_warnings(record=True) as record:
WeightedMetaRegressor(
estimator=LinearRegression().set_fit_request(sample_weight=False)
).fit(X, y, sample_weight=my_weights)
for w in record:
print(w.message)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Received sample_weight of length = 100 in WeightedMetaRegressor.
Support for sample_weight has recently been added to this class. To maintain backward compatibility, it is ignored now. Using `set_fit_request(sample_weight={True, False})` on this method of the class, you can set the request value to False to silence this warning, or to True to consume and use the metadata.
.. GENERATED FROM PYTHON SOURCE LINES 677-679
When an estimator consumes a metadata which it didn't consume before, the
following pattern can be used to warn the users about it.
.. GENERATED FROM PYTHON SOURCE LINES 679-697
.. code-block:: Python
class ExampleRegressor(RegressorMixin, BaseEstimator):
__metadata_request__fit = {"sample_weight": metadata_routing.WARN}
def fit(self, X, y, sample_weight=None):
check_metadata(self, sample_weight=sample_weight)
return self
def predict(self, X):
return np.zeros(shape=(len(X)))
with warnings.catch_warnings(record=True) as record:
MetaRegressor(estimator=ExampleRegressor()).fit(X, y, sample_weight=my_weights)
for w in record:
print(w.message)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
sample_weight is None in ExampleRegressor.
Support for sample_weight has recently been added to this class. To maintain backward compatibility, it is ignored now. Using `set_fit_request(sample_weight={True, False})` on this method of the class, you can set the request value to False to silence this warning, or to True to consume and use the metadata.
.. GENERATED FROM PYTHON SOURCE LINES 698-699
At the end we disable the configuration flag for metadata routing:
.. GENERATED FROM PYTHON SOURCE LINES 699-702
.. code-block:: Python
set_config(enable_metadata_routing=False)
.. GENERATED FROM PYTHON SOURCE LINES 703-720
Third Party Development and scikit-learn Dependency
---------------------------------------------------
As seen above, information is communicated between classes using
:class:`~utils.metadata_routing.MetadataRequest` and
:class:`~utils.metadata_routing.MetadataRouter`. It is strongly not advised,
but possible to vendor the tools related to metadata-routing if you strictly
want to have a scikit-learn compatible estimator, without depending on the
scikit-learn package. If all of the following conditions are met, you do NOT
need to modify your code at all:
- your estimator inherits from :class:`~base.BaseEstimator`
- the parameters consumed by your estimator's methods, e.g. ``fit``, are
explicitly defined in the method's signature, as opposed to being
``*args`` or ``*kwargs``.
- your estimator does not route any metadata to the underlying objects, i.e.
it's not a *router*.
.. rst-class:: sphx-glr-timing
**Total running time of the script:** (0 minutes 0.043 seconds)
.. _sphx_glr_download_auto_examples_miscellaneous_plot_metadata_routing.py:
.. only:: html
.. container:: sphx-glr-footer sphx-glr-footer-example
.. container:: binder-badge
.. image:: images/binder_badge_logo.svg
:target: https://mybinder.org/v2/gh/scikit-learn/scikit-learn/1.6.X?urlpath=lab/tree/notebooks/auto_examples/miscellaneous/plot_metadata_routing.ipynb
:alt: Launch binder
:width: 150 px
.. container:: lite-badge
.. image:: images/jupyterlite_badge_logo.svg
:target: ../../lite/lab/index.html?path=auto_examples/miscellaneous/plot_metadata_routing.ipynb
:alt: Launch JupyterLite
:width: 150 px
.. container:: sphx-glr-download sphx-glr-download-jupyter
:download:`Download Jupyter notebook: plot_metadata_routing.ipynb <plot_metadata_routing.ipynb>`
.. container:: sphx-glr-download sphx-glr-download-python
:download:`Download Python source code: plot_metadata_routing.py <plot_metadata_routing.py>`
.. container:: sphx-glr-download sphx-glr-download-zip
:download:`Download zipped: plot_metadata_routing.zip <plot_metadata_routing.zip>`
.. include:: plot_metadata_routing.recommendations
.. only:: html
.. rst-class:: sphx-glr-signature
`Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_