scikit-learn/sklearn/svm/base.py

488 lines
17 KiB
Python

import numpy as np
from . import libsvm, liblinear
from ..base import BaseEstimator
from ..utils import safe_asanyarray
LIBSVM_IMPL = ['c_svc', 'nu_svc', 'one_class', 'epsilon_svr', 'nu_svr']
def _get_class_weight(class_weight, y):
"""
Estimate class weights for unbalanced datasets.
"""
if class_weight == 'auto':
uy = np.unique(y)
weight_label = np.asarray(uy, dtype=np.int32, order='C')
weight = np.array([1.0 / np.sum(y == i) for i in uy],
dtype=np.float64, order='C')
weight *= uy.shape[0] / np.sum(weight)
else:
if class_weight is None:
keys = values = []
else:
keys = class_weight.keys()
values = class_weight.values()
weight = np.asarray(values, dtype=np.float64, order='C')
weight_label = np.asarray(keys, dtype=np.int32, order='C')
return weight, weight_label
class BaseLibSVM(BaseEstimator):
"""
Base class for classifiers that use libsvm as library for
support vector machine classification and regression.
Should not be used directly, use derived classes instead
"""
def __init__(self, impl, kernel, degree, gamma, coef0,
tol, C, nu, epsilon, shrinking, probability):
if not impl in LIBSVM_IMPL:
raise ValueError("impl should be one of %s, %s was given" % (
LIBSVM_IMPL, impl))
if hasattr(kernel, '__call__'):
self.kernel_function = kernel
self.kernel = 'precomputed'
else:
self.kernel = kernel
self.impl = impl
self.degree = degree
self.gamma = gamma
self.coef0 = coef0
self.tol = tol
self.C = C
self.nu = nu
self.epsilon = epsilon
self.shrinking = shrinking
self.probability = probability
def _compute_kernel(self, X):
""" Return the data transformed by the kernel (if the kernel
is a callable).
"""
if hasattr(self, 'kernel_function'):
# in the case of precomputed kernel given as a function, we
# have to compute explicitly the kernel matrix
X = np.asanyarray(self.kernel_function(X, self.__Xfit),
dtype=np.float64, order='C')
return X
def fit(self, X, y, class_weight=None, sample_weight=None, cache_size=100.):
"""
Fit the SVM model according to the given training data and
parameters.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Training vectors, where n_samples is the number of samples
and n_features is the number of features.
y : array-like, shape = [n_samples]
Target values (integers in classification, real numbers in
regression)
class_weight : {dict, 'auto'}, optional
Set the parameter C of class i to class_weight[i]*C for
SVC. If not given, all classes are supposed to have
weight one. The 'auto' mode uses the values of y to
automatically adjust weights inversely proportional to
class frequencies.
sample_weight : array-like, shape = [n_samples], optional
Weights applied to individual samples (1. for unweighted).
cache_size: float, optional
Specify the size of the cache (in MB)
Returns
-------
self : object
Returns self.
Notes
------
If X and y are not C-ordered and contiguous arrays, they are
copied.
"""
X = np.asanyarray(X, dtype=np.float64, order='C')
y = np.asanyarray(y, dtype=np.float64, order='C')
sample_weight = np.asanyarray([] if sample_weight is None
else sample_weight, dtype=np.float64)
if hasattr(self, 'kernel_function'):
# you must store a reference to X to compute the kernel in predict
# TODO: add keyword copy to copy on demand
self.__Xfit = X
X = self._compute_kernel(X)
class_weight, class_weight_label = \
_get_class_weight(class_weight, y)
# check dimensions
solver_type = LIBSVM_IMPL.index(self.impl)
if solver_type != 2 and X.shape[0] != y.shape[0]:
raise ValueError("X and y have incompatible shapes.\n" +
"X has %s features, but y has %s." % \
(X.shape[0], y.shape[0]))
if self.kernel == "precomputed" and X.shape[0] != X.shape[1]:
raise ValueError("X.shape[0] should be equal to X.shape[1]")
if (self.kernel in ['poly', 'rbf']) and (self.gamma == 0):
# if custom gamma is not provided ...
self.gamma = 1.0 / X.shape[0]
self.shape_fit_ = X.shape
self.support_, self.support_vectors_, self.n_support_, \
self.dual_coef_, self.intercept_, self.label_, self.probA_, \
self.probB_ = libsvm.fit(X, y,
svm_type=solver_type, sample_weight=sample_weight,
class_weight=class_weight,
class_weight_label=class_weight_label,
**self._get_params())
return self
def predict(self, X):
"""
This function does classification or regression on an array of
test vectors X.
For a classification model, the predicted class for each
sample in X is returned. For a regression model, the function
value of X calculated is returned.
For an one-class model, +1 or -1 is returned.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Returns
-------
C : array, shape = [n_samples]
"""
X = np.asanyarray(X, dtype=np.float64, order='C')
if X.ndim == 1:
# don't use np.atleast_2d, it doesn't guarantee C-contiguity
X = np.reshape(X, (1, -1), order='C')
n_samples, n_features = X.shape
X = self._compute_kernel(X)
if self.kernel == "precomputed":
if X.shape[1] != self.shape_fit_[0]:
raise ValueError("X.shape[1] should be equal to the number of "
"samples at training time!")
elif n_features != self.shape_fit_[1]:
raise ValueError("X.shape[1] should be equal to the number of "
"features at training time!")
svm_type = LIBSVM_IMPL.index(self.impl)
return libsvm.predict(
X, self.support_, self.support_vectors_, self.n_support_,
self.dual_coef_, self.intercept_,
self.label_, self.probA_, self.probB_,
svm_type=svm_type, **self._get_params())
def predict_proba(self, X):
"""
This function does classification or regression on a test vector X
given a model with probability information.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Returns
-------
X : array-like, shape = [n_samples, n_classes]
Returns the probability of the sample for each class in
the model, where classes are ordered by arithmetical
order.
Notes
-----
The probability model is created using cross validation, so
the results can be slightly different than those obtained by
predict. Also, it will meaningless results on very small
datasets.
"""
if not self.probability:
raise ValueError(
"probability estimates must be enabled to use this method")
X = np.asanyarray(X, dtype=np.float64, order='C')
if X.ndim == 1:
# don't use np.atleast_2d, it doesn't guarantee C-contiguity
X = np.reshape(X, (1, -1), order='C')
X = self._compute_kernel(X)
if self.impl not in ('c_svc', 'nu_svc'):
raise NotImplementedError("predict_proba only implemented for SVC "
"and NuSVC")
svm_type = LIBSVM_IMPL.index(self.impl)
pprob = libsvm.predict_proba(
X, self.support_, self.support_vectors_, self.n_support_,
self.dual_coef_, self.intercept_, self.label_,
self.probA_, self.probB_,
svm_type=svm_type, **self._get_params())
return pprob
def predict_log_proba(self, T):
"""
This function does classification or regression on a test vector T
given a model with probability information.
Parameters
----------
T : array-like, shape = [n_samples, n_features]
Returns
-------
T : array-like, shape = [n_samples, n_classes]
Returns the log-probabilities of the sample for each class in
the model, where classes are ordered by arithmetical
order.
Notes
-----
The probability model is created using cross validation, so
the results can be slightly different than those obtained by
predict. Also, it will meaningless results on very small
datasets.
"""
return np.log(self.predict_proba(T))
def decision_function(self, X):
"""
Calculate the distance of the samples T to the separating hyperplane.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Returns
-------
X : array-like, shape = [n_samples, n_class * (n_class-1) / 2]
Returns the decision function of the sample for each class
in the model.
"""
X = np.asanyarray(X, dtype=np.float64, order='C')
if X.ndim == 1:
# don't use np.atleast_2d, it doesn't guarantee C-contiguity
X = np.reshape(X, (1, -1), order='C')
X = self._compute_kernel(X)
dec_func = libsvm.decision_function(
X, self.support_, self.support_vectors_, self.n_support_,
self.dual_coef_, self.intercept_, self.label_,
self.probA_, self.probB_,
svm_type=LIBSVM_IMPL.index(self.impl),
**self._get_params())
if self.impl != 'one_class':
# libsvm has the convention of returning negative values for
# rightmost labels, so we invert the sign since our label_ is
# sorted by increasing order
return -dec_func
else:
return dec_func
@property
def coef_(self):
if self.kernel != 'linear':
raise NotImplementedError('coef_ is only available when using a '
'linear kernel')
return np.dot(self.dual_coef_, self.support_vectors_)
class BaseLibLinear(BaseEstimator):
"""
Base for classes binding liblinear (dense and sparse versions)
"""
_solver_type_dict = {
'PL2_LLR_D0' : 0, # L2 penalty, logistic regression
'PL2_LL2_D1' : 1, # L2 penalty, L2 loss, dual form
'PL2_LL2_D0' : 2, # L2 penalty, L2 loss, primal form
'PL2_LL1_D1' : 3, # L2 penalty, L1 Loss, dual form
'MC_SVC' : 4, # Multi-class Support Vector Classification
'PL1_LL2_D0' : 5, # L1 penalty, L2 Loss, primal form
'PL1_LLR_D0' : 6, # L1 penalty, logistic regression
'PL2_LLR_D1' : 7, # L2 penalty, logistic regression, dual form
}
def __init__(self, penalty='l2', loss='l2', dual=True, tol=1e-4, C=1.0,
multi_class=False, fit_intercept=True, intercept_scaling=1):
self.penalty = penalty
self.loss = loss
self.dual = dual
self.tol = tol
self.C = C
self.fit_intercept = fit_intercept
self.intercept_scaling = intercept_scaling
self.multi_class = multi_class
# Check that the arguments given are valid:
self._get_solver_type()
def _get_solver_type(self):
""" Return the magic number for the solver described by the
settings.
"""
if self.multi_class:
solver_type = 'MC_SVC'
else:
solver_type = "P%s_L%s_D%d" % (
self.penalty.upper(), self.loss.upper(), int(self.dual))
if not solver_type in self._solver_type_dict:
raise ValueError('Not supported set of arguments: '
+ solver_type)
return self._solver_type_dict[solver_type]
def fit(self, X, y, class_weight=None):
"""
Fit the model according to the given training data and
parameters.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Training vector, where n_samples in the number of samples and
n_features is the number of features.
y : array-like, shape = [n_samples]
Target vector relative to X
class_weight : {dict, 'auto'}, optional
Weights associated with classes. If not given, all classes
are supposed to have weight one.
Returns
-------
self : object
Returns self.
"""
self.class_weight, self.class_weight_label = \
_get_class_weight(class_weight, y)
X = safe_asanyarray(X, dtype=np.float64, order='C')
if not isinstance(X, np.ndarray): # sparse X passed in by user
raise ValueError("Training vectors should be array-like, not %s"
% type(X))
y = np.asanyarray(y, dtype=np.int32, order='C')
self.raw_coef_, self.label_ = liblinear.train_wrap(X, y,
self._get_solver_type(), self.tol,
self._get_bias(), self.C,
self.class_weight_label, self.class_weight)
return self
def predict(self, X):
"""
Predict target values of X according to the fitted model.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Returns
-------
C : array, shape = [n_samples]
"""
X = np.asanyarray(X, dtype=np.float64, order='C')
self._check_n_features(X)
coef = self.raw_coef_
return liblinear.predict_wrap(X, coef,
self._get_solver_type(),
self.tol, self.C,
self.class_weight_label,
self.class_weight, self.label_,
self._get_bias())
def decision_function(self, X):
"""
Return the decision function of X according to the trained
model.
Parameters
----------
X : array-like, shape = [n_samples, n_features]
Returns
-------
T : array-like, shape = [n_samples, n_class]
Returns the decision function of the sample for each class
in the model.
"""
X = np.asanyarray(X, dtype=np.float64, order='C')
if X.ndim == 1:
# don't use np.atleast_2d, it doesn't guarantee C-contiguity
X = np.reshape(X, (1, -1), order='C')
self._check_n_features(X)
dec_func = liblinear.decision_function_wrap(
X, self.raw_coef_, self._get_solver_type(), self.tol,
self.C, self.class_weight_label, self.class_weight,
self.label_, self._get_bias())
if len(self.label_) <= 2:
# in the two-class case, the decision sign needs be flipped
# due to liblinear's design
return -dec_func
else:
return dec_func
def _check_n_features(self, X):
n_features = self.raw_coef_.shape[1]
if self.fit_intercept:
n_features -= 1
if X.shape[1] != n_features:
raise ValueError("X.shape[1] should be %d, not %d." % (n_features,
X.shape[1]))
@property
def intercept_(self):
if self.fit_intercept:
ret = self.intercept_scaling * self.raw_coef_[:, -1]
if len(self.label_) <= 2:
ret *= -1
return ret
return 0.0
@property
def coef_(self):
if self.fit_intercept:
ret = self.raw_coef_[:, : -1]
else:
ret = self.raw_coef_
if len(self.label_) <= 2:
return -ret
else:
return ret
def predict_proba(self, T):
# only available for logistic regression
raise NotImplementedError(
'liblinear does not provide this functionality')
def _get_bias(self):
if self.fit_intercept:
return self.intercept_scaling
else:
return -1.0
libsvm.set_verbosity_wrap(0)