Skip to content

[MRG+2] add get_feature_names to PolynomialFeatures #6372

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Feb 25, 2016
Merged

[MRG+2] add get_feature_names to PolynomialFeatures #6372

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
3fa684d
add get_feature_names to PolynomialFeatures
amueller Feb 19, 2016
ddc1740
fix PolynomialFeatures.powers_ in python 0.16.1
amueller Feb 19, 2016
897f8b6
don't do ^1
amueller Feb 22, 2016
8fb928d
fixed doc for powers, added test
amueller Feb 24, 2016
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
35 changes: 33 additions & 2 deletions sklearn/preprocessing/data.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
from ..utils.extmath import row_norms
from ..utils.extmath import _incremental_mean_and_var
from ..utils.fixes import combinations_with_replacement as combinations_w_r
from ..utils.fixes import bincount
from ..utils.sparsefuncs_fast import (inplace_csr_row_normalize_l1,
inplace_csr_row_normalize_l2)
from ..utils.sparsefuncs import (inplace_column_scale,
Expand Down Expand Up @@ -1140,7 +1141,7 @@ class PolynomialFeatures(BaseEstimator, TransformerMixin):

Attributes
----------
powers_ : array, shape (n_input_features, n_output_features)
powers_ : array, shape (n_output_features, n_input_features)
powers_[i, j] is the exponent of the jth input in the ith output.

n_input_features_ : int
Expand Down Expand Up @@ -1179,9 +1180,39 @@ def powers_(self):
combinations = self._combinations(self.n_input_features_, self.degree,
self.interaction_only,
self.include_bias)
return np.vstack(np.bincount(c, minlength=self.n_input_features_)
return np.vstack(bincount(c, minlength=self.n_input_features_)
for c in combinations)

def get_feature_names(self, input_features=None):
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've been suggesting for a while that this is an appropriate way to deal with feature names in extractor/transformer pipelines. I'm happy to see this.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1. Very nice.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've been wanting to do it but didn't have time. I'm just writing the book chapter about preprocessing and I'm embarrassed not having that feature ;)

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hey, who knew writing books would be good for something?

"""
Return feature names for output features

Parameters
----------
input_features : list of string, length n_features, optional
String names for input features if available. By default,
"x0", "x1", ... "xn_features" is used.

Returns
-------
output_feature_names : list of string, length n_output_features
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just a nitpick here:
Maybe there should be a "." in the end of this line?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

conventionally not.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh I see.
So this line:
https://github.com/amueller/scikit-learn/blob/poly_feature_names/sklearn/preprocessing/data.py#L1586
ends with a "." because it's a new sentence used to describe output?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, it is description, not type.


"""
powers = self.powers_
if input_features is None:
input_features = ['x%d' % i for i in range(powers.shape[1])]
feature_names = []
for row in powers:
inds = np.where(row)[0]
if len(inds):
name = " ".join("%s^%d" % (input_features[ind], exp)
if exp != 1 else input_features[ind]
for ind, exp in zip(inds, row[inds]))
else:
name = "1"
feature_names.append(name)
return feature_names

def fit(self, X, y=None):
"""
Compute number of output features.
Expand Down
25 changes: 25 additions & 0 deletions sklearn/preprocessing/tests/test_data.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,7 @@
import numpy.linalg as la
from scipy import sparse
from distutils.version import LooseVersion
from sklearn.externals.six import u

from sklearn.utils import gen_batches

Expand Down Expand Up @@ -118,6 +119,30 @@ def test_polynomial_features():
X_poly = interact.fit_transform(X)
assert_array_almost_equal(X_poly, P2[:, [0, 1, 2, 4]])

assert_equal(interact.powers_.shape, (interact.n_output_features_,
interact.n_input_features_))


def test_polynomial_feature_names():
X = np.arange(30).reshape(10, 3)
poly = PolynomialFeatures(degree=2, include_bias=True).fit(X)
feature_names = poly.get_feature_names()
assert_array_equal(['1', 'x0', 'x1', 'x2', 'x0^2', 'x0 x1',
'x0 x2', 'x1^2', 'x1 x2', 'x2^2'],
feature_names)

poly = PolynomialFeatures(degree=3, include_bias=False).fit(X)
feature_names = poly.get_feature_names(["a", "b", "c"])
assert_array_equal(['a', 'b', 'c', 'a^2', 'a b', 'a c', 'b^2',
'b c', 'c^2', 'a^3', 'a^2 b', 'a^2 c',
'a b^2', 'a b c', 'a c^2', 'b^3', 'b^2 c',
'b c^2', 'c^3'], feature_names)
# test some unicode
poly = PolynomialFeatures(degree=1, include_bias=True).fit(X)
feature_names = poly.get_feature_names([u"\u0001F40D", u"\u262E", u"\u05D0"])
assert_array_equal([u"1", u"\u0001F40D", u"\u262E", u"\u05D0"],
feature_names)


def test_standard_scaler_1d():
# Test scaling of dataset along single axis
Expand Down