diff options
| author | cheusov <cheusov@pkgsrc.org> | 2014-10-19 09:57:21 +0000 |
|---|---|---|
| committer | cheusov <cheusov@pkgsrc.org> | 2014-10-19 09:57:21 +0000 |
| commit | 28890509d95c522503728fbd85b717c6081f2621 (patch) | |
| tree | b68e350257abaf0da275a3c3dd5cfd20e3adf5a0 /math/liblinear | |
| parent | 50d7b1f0a83eae61efd5a2835a86ab0adf319f7b (diff) | |
| download | pkgsrc-28890509d95c522503728fbd85b717c6081f2621.tar.gz | |
Add liblinear.
LIBLINEAR is a linear classifier for data with millions of instances
and features. It supports
L2-regularized classifiers
L2-loss linear SVM, L1-loss linear SVM, and logistic regression (LR)
L1-regularized classifiers (after version 1.4)
L2-loss linear SVM and logistic regression (LR)
L2-regularized support vector regression (after version 1.9)
L2-loss linear SVR and L1-loss linear SVR.
Main features of LIBLINEAR include
Same data format as LIBSVM, our general-purpose SVM solver,
and also similar usage
Multi-class classification: 1) one-vs-the rest, 2) Crammer & Singer
Cross validation for model selection
Probability estimates (logistic regression only)
Weights for unbalanced data
MATLAB/Octave, Java, Python, Ruby interfaces
Diffstat (limited to 'math/liblinear')
| -rw-r--r-- | math/liblinear/DESCR | 16 | ||||
| -rw-r--r-- | math/liblinear/Makefile | 28 | ||||
| -rw-r--r-- | math/liblinear/PLIST | 8 | ||||
| -rw-r--r-- | math/liblinear/distinfo | 5 |
4 files changed, 57 insertions, 0 deletions
diff --git a/math/liblinear/DESCR b/math/liblinear/DESCR new file mode 100644 index 00000000000..b91e706c498 --- /dev/null +++ b/math/liblinear/DESCR @@ -0,0 +1,16 @@ +LIBLINEAR is a linear classifier for data with millions of instances +and features. It supports + L2-regularized classifiers + L2-loss linear SVM, L1-loss linear SVM, and logistic regression (LR) + L1-regularized classifiers (after version 1.4) + L2-loss linear SVM and logistic regression (LR) + L2-regularized support vector regression (after version 1.9) + L2-loss linear SVR and L1-loss linear SVR. +Main features of LIBLINEAR include + Same data format as LIBSVM, our general-purpose SVM solver, + and also similar usage + Multi-class classification: 1) one-vs-the rest, 2) Crammer & Singer + Cross validation for model selection + Probability estimates (logistic regression only) + Weights for unbalanced data + MATLAB/Octave, Java, Python, Ruby interfaces diff --git a/math/liblinear/Makefile b/math/liblinear/Makefile new file mode 100644 index 00000000000..a5dfcdb890b --- /dev/null +++ b/math/liblinear/Makefile @@ -0,0 +1,28 @@ +# $NetBSD: Makefile,v 1.1.1.1 2014/10/19 09:57:21 cheusov Exp $ + +DISTNAME= liblinear-1.94 +CATEGORIES= math +MASTER_SITES= http://www.csie.ntu.edu.tw/~cjlin/liblinear/ + +MAINTAINER= cheusov@NetBSD.org +HOMEPAGE= http://www.csie.ntu.edu.tw/~cjlin/liblinear/ +COMMENT= Library for large linear classification +LICENSE= modified-bsd + +USE_LANGUAGES= c c++ + +BUILD_TARGET= all lib +AUTO_MKDIRS= yes + +do-install: + ${INSTALL_PROGRAM} ${WRKSRC}/predict ${DESTDIR}${PREFIX}/bin/liblinear-predict + ${INSTALL_PROGRAM} ${WRKSRC}/train ${DESTDIR}${PREFIX}/bin/liblinear-train + ${INSTALL_LIB} ${WRKSRC}/liblinear.so.1 ${DESTDIR}${PREFIX}/lib + ${LN} -s liblinear.so.1 ${DESTDIR}${PREFIX}/lib/liblinear.so + ${INSTALL_DATA} ${WRKSRC}/linear.h ${DESTDIR}${PREFIX}/include + ${INSTALL_DATA} ${WRKSRC}/README \ + ${DESTDIR}${PREFIX}/share/doc/${PKGBASE} + ${INSTALL_DATA} ${WRKSRC}/heart_scale \ + ${DESTDIR}${PREFIX}/share/examples/${PKGBASE} + +.include "../../mk/bsd.pkg.mk" diff --git a/math/liblinear/PLIST b/math/liblinear/PLIST new file mode 100644 index 00000000000..c0dc24e1ca6 --- /dev/null +++ b/math/liblinear/PLIST @@ -0,0 +1,8 @@ +@comment $NetBSD: PLIST,v 1.1.1.1 2014/10/19 09:57:21 cheusov Exp $ +bin/liblinear-predict +bin/liblinear-train +include/linear.h +lib/liblinear.so +lib/liblinear.so.1 +share/doc/liblinear/README +share/examples/liblinear/heart_scale diff --git a/math/liblinear/distinfo b/math/liblinear/distinfo new file mode 100644 index 00000000000..149911c005a --- /dev/null +++ b/math/liblinear/distinfo @@ -0,0 +1,5 @@ +$NetBSD: distinfo,v 1.1.1.1 2014/10/19 09:57:21 cheusov Exp $ + +SHA1 (liblinear-1.94.tar.gz) = 19678355e6c933b7ec133e07fef77796e50df0d5 +RMD160 (liblinear-1.94.tar.gz) = 56cc112d3e4397804d67eea4e266b363c822307b +Size (liblinear-1.94.tar.gz) = 335400 bytes |