Skip to content

Commit

Permalink
Move to sklearn 0.20's OneHotEncoder and fix README results
Browse files Browse the repository at this point in the history
  • Loading branch information
@MaxHalford
MaxHalford committed Oct 11, 2018
1 parent 5734562 commit fb9eff1
Show file tree
Hide file tree
Showing 6 changed files with 47 additions and 165 deletions.
53 changes: 32 additions & 21 deletions README.md
View file
Expand Up @@ -29,9 +29,20 @@

<br/>

## Introduction
Prince is a library for doing [factor analysis](https://www.wikiwand.com/en/Factor_analysis). This includes a variety of methods including [principal component analysis (PCA)](https://www.wikiwand.com/en/Principal_component_analysis) and [correspondence analysis (CA)](https://www.wikiwand.com/en/Correspondence_analysis). The goal is to provide an efficient implementation for each algorithm along with a scikit-learn API.

Prince is a library for doing [factor analysis](https://www.wikiwand.com/en/Factor_analysis). This includes a variety of methods including [principal component analysis (PCA)](https://www.wikiwand.com/en/Principal_component_analysis) and [correspondence analysis (CA)](https://www.wikiwand.com/en/Correspondence_analysis). The goal is to provide an efficient implementation for each algorithm along with a nice API.
## Table of contents

- [Installation](#installation)
- [Usage](#usage)
- [Guidelines](#guidelines)
- [Principal component analysis (PCA)](#principal-component-analysis-(pca))
- [Correspondence analysis (CA)](#correspondence-analysis-(ca))
- [Multiple correspondence analysis (MCA)](#multiple-correspondence-analysis-(mca))
- [Multiple factor analysis (MFA)](#multiple-factor-analysis-(mfa))
- [Factor analysis of mixed data (FAMD)](#factor-analysis-of-mixed-data-(famd))
- [Going faster](#going-faster)
- [License](#license)

## Installation

Expand Down Expand Up @@ -137,11 +148,11 @@ Once the `PCA` has been fitted, it can be used to extract the row principal coor
```python
>>> pca.transform(X).head() # Same as pca.row_coordinates(X).head()
0 1
0 -2.264542 0.505704
1 -2.086426 -0.655405
2 -2.367950 -0.318477
3 -2.304197 -0.575368
4 -2.388777 0.674767
0 -2.264703 0.480027
1 -2.080961 -0.674134
2 -2.364229 -0.341908
3 -2.299384 -0.597395
4 -2.389842 0.646835

```

Expand Down Expand Up @@ -172,21 +183,21 @@ Each principal component explains part of the underlying of the distribution. Yo

```python
>>> pca.explained_inertia_ # doctest: +ELLIPSIS
[0.727704..., 0.230305...]
[0.729624..., 0.228507...]

```

The explained inertia represents the percentage of the inertia each principal component contributes. It sums up to 1 if the `n_components` property is equal to the number of columns in the original dataset. you The explained inertia is obtained by dividing the eigenvalues obtained with the SVD by the total inertia, both of which are also accessible.

```python
>>> pca.eigenvalues_ # doctest: +ELLIPSIS
[436.622712..., 138.183139...]
[437.774672..., 137.104570...]

>>> pca.total_inertia_
600.0
>>> pca.total_inertia_ # doctest: +ELLIPSIS
600.0...

>>> pca.explained_inertia_
[0.727704..., 0.230305...]
[0.729624..., 0.228507...]

```

Expand All @@ -195,10 +206,10 @@ You can also obtain the correlations between the original variables and the prin
```python
>>> pca.column_correlations(X)
0 1
Petal length 0.991684 0.020247
Petal width 0.964996 0.062786
Sepal length 0.891224 0.357352
Sepal width -0.449313 0.888351
Petal length 0.991555 0.023415
Petal width 0.964979 0.064000
Sepal length 0.890169 0.360830
Sepal width -0.460143 0.882716

```

Expand All @@ -207,11 +218,11 @@ You may also want to know how much each observation contributes to each principa
```python
>>> pca.row_contributions(X).head()
0 1
0 0.011745 0.001851
1 0.009970 0.003109
2 0.012842 0.000734
3 0.012160 0.002396
4 0.013069 0.003295
0 0.011716 0.001681
1 0.009892 0.003315
2 0.012768 0.000853
3 0.012077 0.002603
4 0.013046 0.003052

```

Expand Down
Binary file modified images/pca_row_coordinates.png
View file
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
1 change: 1 addition & 0 deletions prince/mca.py
View file
Expand Up @@ -2,6 +2,7 @@
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from sklearn import preprocessing
from sklearn import utils

from . import ca
Expand Down
146 changes: 8 additions & 138 deletions prince/one_hot.py
View file
@@ -1,150 +1,18 @@
"""This module contains a custom one-hot encoder. It inherits from sklearn's
OneHotEncoder and returns a pandas DataFrame with appropriate column names.
OneHotEncoder and returns a pandas.SparseDataFrame with appropriate column
names and index values.
"""
import itertools

import numpy as np
import pandas as pd
from scipy import sparse
from sklearn import preprocessing
from sklearn import base
from sklearn import utils


class CategoricalEncoder(base.BaseEstimator, base.TransformerMixin):

def __init__(self, encoding='onehot', categories='auto', dtype=np.float64,
handle_unknown='error'):
self.encoding = encoding
self.categories = categories
self.dtype = dtype
self.handle_unknown = handle_unknown

def fit(self, X, y=None):
"""Fit the CategoricalEncoder to X.
Parameters
----------
X : array-like, shape [n_samples, n_features]
The data to determine the categories of each feature.
Returns
-------
self
"""
if self.encoding not in ['onehot', 'onehot-dense', 'ordinal']:
template = ("encoding should be either 'onehot', 'onehot-dense' "
"or 'ordinal', got %s")
raise ValueError(template % self.handle_unknown)

if self.handle_unknown not in ['error', 'ignore']:
template = ("handle_unknown should be either 'error' or "
"'ignore', got %s")
raise ValueError(template % self.handle_unknown)

if self.encoding == 'ordinal' and self.handle_unknown == 'ignore':
raise ValueError("handle_unknown='ignore' is not supported for"
" encoding='ordinal'")

if self.categories != 'auto':
for cats in self.categories:
if not np.all(np.sort(cats) == np.array(cats)):
raise ValueError("Unsorted categories are not yet "
"supported")

X_temp = utils.check_array(X, dtype=None)
if not hasattr(X, 'dtype') and np.issubdtype(X_temp.dtype, np.str_):
X = utils.check_array(X, dtype=np.object)
else:
X = X_temp

n_samples, n_features = X.shape

self._label_encoders_ = [preprocessing.LabelEncoder() for _ in range(n_features)]

for i in range(n_features):
le = self._label_encoders_[i]
Xi = X[:, i]
if self.categories == 'auto':
le.fit(Xi)
else:
if self.handle_unknown == 'error':
valid_mask = np.in1d(Xi, self.categories[i])
if not np.all(valid_mask):
diff = np.unique(Xi[~valid_mask])
msg = ("Found unknown categories {0} in column {1}"
" during fit".format(diff, i))
raise ValueError(msg)
le.classes_ = np.array(self.categories[i])

self.categories_ = [le.classes_ for le in self._label_encoders_]

return self

def transform(self, X):
"""Transform X using specified encoding scheme.
Parameters
----------
X : array-like, shape [n_samples, n_features]
The data to encode.
Returns
-------
X_out : sparse matrix or a 2-d array
Transformed input.
"""
X_temp = utils.check_array(X, dtype=None)
if not hasattr(X, 'dtype') and np.issubdtype(X_temp.dtype, np.str_):
X = utils.check_array(X, dtype=np.object)
else:
X = X_temp

n_samples, n_features = X.shape
X_int = np.zeros_like(X, dtype=np.int)
X_mask = np.ones_like(X, dtype=np.bool)

for i in range(n_features):
Xi = X[:, i]
valid_mask = np.in1d(Xi, self.categories_[i])

if not np.all(valid_mask):
if self.handle_unknown == 'error':
diff = np.unique(X[~valid_mask, i])
msg = ("Found unknown categories {0} in column {1}"
" during transform".format(diff, i))
raise ValueError(msg)
else:
# Set the problematic rows to an acceptable value and
# continue `The rows are marked `X_mask` and will be
# removed later.
X_mask[:, i] = valid_mask
Xi = Xi.copy()
Xi[~valid_mask] = self.categories_[i][0]
X_int[:, i] = self._label_encoders_[i].transform(Xi)

if self.encoding == 'ordinal':
return X_int.astype(self.dtype, copy=False)

mask = X_mask.ravel()
n_values = [cats.shape[0] for cats in self.categories_]
n_values = np.array([0] + n_values)
feature_indices = np.cumsum(n_values)

indices = (X_int + feature_indices[:-1]).ravel()[mask]
indptr = X_mask.sum(axis=1).cumsum()
indptr = np.insert(indptr, 0, 0)
data = np.ones(n_samples * n_features)[mask]

out = sparse.csr_matrix((data, indices, indptr),
shape=(n_samples, feature_indices[-1]),
dtype=self.dtype)
if self.encoding == 'onehot-dense':
return out.toarray()
else:
return out


class OneHotEncoder(CategoricalEncoder):
class OneHotEncoder(preprocessing.OneHotEncoder):

def __init__(self):
super().__init__(encoding='onehot-dense')
super().__init__(sparse=True, dtype=np.uint8)

def fit(self, X, y=None):

Expand All @@ -159,11 +27,13 @@ def fit(self, X, y=None):
]
for i, col in enumerate(X.columns)
]))

return self

def transform(self, X):
return pd.DataFrame(
return pd.SparseDataFrame(
data=super().transform(X),
columns=self.column_names_,
index=X.index if isinstance(X, pd.DataFrame) else None
index=X.index if isinstance(X, pd.DataFrame) else None,
default_fill_value=0
)
8 changes: 4 additions & 4 deletions requirements.dev.txt
View file
@@ -1,10 +1,10 @@
coveralls>=1.3.0
nose>=1.3.7
pytest>=3.5.0
pytest-cov>=2.5.1
pytest>=3.5.1
pytest-cov>=2.6.0
fbpca>=1.0
matplotlib>=2.2.2
numpy>=1.14.0
pandas>=0.22.0
pandas>=0.23.0
scipy>=1.0.1
scikit-learn>=0.19.1
scikit-learn>=0.20.0
4 changes: 2 additions & 2 deletions setup.py
View file
Expand Up @@ -25,9 +25,9 @@
REQUIRED = [
'matplotlib>=2.2.2',
'numpy>=1.14.0',
'pandas>=0.22.0',
'pandas>=0.23.0',
'scipy>=1.0.1',
'scikit-learn>=0.19.1'
'scikit-learn>=0.20.0'
]

# The rest you shouldn't have to touch too much :)
Expand Down

0 comments on commit fb9eff1

Please sign in to comment.