From 9d0fc56170381d72501f865467041b9d6fa696c2 Mon Sep 17 00:00:00 2001
From: Fabian Pedregosa <fabian.pedregosa@inria.fr>
Date: Mon, 3 May 2010 09:25:21 +0000
Subject: [PATCH] ENH: add docstrings to module svm
git-svn-id: https://scikit-learn.svn.sourceforge.net/svnroot/scikit-learn/trunk@722 22fbfee3-77ab-4535-9bad-27d1bd3bc7d8
---
scikits/learn/svm.py | 61 ++++++++++++++++++++++++++++++++++++++++----
1 file changed, 56 insertions(+), 5 deletions(-)
diff --git a/scikits/learn/svm.py b/scikits/learn/svm.py
index 849a2c2b99..9d65748418 100644
--- a/scikits/learn/svm.py
+++ b/scikits/learn/svm.py
@@ -185,8 +185,14 @@ class SVC(BaseLibsvm):
`support_` : array-like, shape = [nSV, nfeatures]
Support vectors
- `dual_coef_` : array, shape = [nclasses-1, nfeatures]
- Coefficient of the support vector in the decision function.
+ `dual_coef_` : array, shape = [nclasses-1, nSV]
+ Coefficient of the support vector in the decision function,
+ where nclasses is the number of classes and nSV is the number
+ of support vectors.
+
+ `coef_` : array, shape = [nclasses-1, nfeatures]
+ Wiehgiths asigned to the features (coefficients in the primal
+ problem). This is only available in the case of linear kernel.
`intercept_` : array, shape = [nclasses-1]
constants in decision function
@@ -230,8 +236,14 @@ class SVR(BaseLibsvm):
`support_` : array-like, shape = [nSV, nfeatures]
Support vectors
- `dual_coef_` : array, shape = [nclasses-1, nfeatures]
- Coefficient of the support vector in the decision function.
+ `dual_coef_` : array, shape = [nclasses-1, nSV]
+ Coefficient of the support vector in the decision function,
+ where nclasses is the number of classes and nSV is the number
+ of support vectors.
+
+ `coef_` : array, shape = [nclasses-1, nfeatures]
+ Wiehgiths asigned to the features (coefficients in the primal
+ problem). This is only available in the case of linear kernel.
`intercept_` : array, shape = [nclasses-1]
constants in decision function
@@ -259,6 +271,23 @@ class OneClassSVM(BaseLibsvm):
"""
Outlayer detection
+ Attributes
+ ----------
+ `support_` : array-like, shape = [nSV, nfeatures]
+ Support vectors
+
+ `dual_coef_` : array, shape = [nclasses-1, nSV]
+ Coefficient of the support vector in the decision function,
+ where nclasses is the number of classes and nSV is the number
+ of support vectors.
+
+ `coef_` : array, shape = [nclasses-1, nfeatures]
+ Wiehgiths asigned to the features (coefficients in the primal
+ problem). This is only available in the case of linear kernel.
+
+ `intercept_` : array, shape = [nclasses-1]
+ constants in decision function
+
Methods
-------
fit(X, Y) : self
@@ -307,7 +336,29 @@ class LinearSVC(object):
penalty is the standard used in SVC. The 'l1' leads to coef_
vectors that are sparse.
- TODO: wrap Cramer & Singer
+
+ Attributes
+ ----------
+ `support_` : array-like, shape = [nSV, nfeatures]
+ Support vectors
+
+ `dual_coef_` : array, shape = [nclasses-1, nSV]
+ Coefficient of the support vector in the decision function,
+ where nclasses is the number of classes and nSV is the number
+ of support vectors.
+
+ `coef_` : array, shape = [nclasses-1, nfeatures]
+ Wiehgiths asigned to the features (coefficients in the primal
+ problem). This is only available in the case of linear kernel.
+
+ `intercept_` : array, shape = [nclasses-1]
+ constants in decision function
+
+
+ Notes
+ -----
+ Some features of liblinear are still not wrapped, like the Cramer
+ & Singer algorithm.
References
----------
--
GitLab