本文首发于个人博客https://kezunlin.me/post/1cd6a04d/，欢迎阅读最新内容！

tutorial to compile and install mplack on ubuntu 16.04
<!--more-->

Guide

mlpack: a scalable C++ machine learning library

dependencies

• Armadillo >= 6.500.0
• Boost
• CMake >= 3.3.2

Armadillo: c++ linear algebra library based on LAPACK and BLAS
If you are compiling Armadillo by hand, ensure that LAPACK and BLAS are enabled.

see OpenCV vs. Armadillo vs. Eigen on Linux

sudo apt-get install libarmadillo-dev

install

apt-get

sudo apt-get install libmlpack-dev

version: 2.0.1
by default mlpack will install to /usr/include/mlpack and /usr/lib

compile

wget https://www.mlpack.org/files/mlpack-3.1.1.tar.gz
git clone https://github.com/mlpack/mlpack.git
mkdir build && cd build && cmake-gui ..
make -j8
sudo make install

configure and output

...
Found Armadillo: /usr/lib/libarmadillo.so (found suitable version "6.500.5", minimum required is "6.500.0") 
Armadillo libraries: /usr/lib/libarmadillo.so
...

version: 3.1.1
by default mlpack will install to /usr/local/include and /usr/local/lib/libmlpack.so.3.1

<script async src="https://pagead2.googlesyndication.com/pagead/js/adsbygoogle.js"></script>
<!-- kzl in-article ad -->
<ins class="adsbygoogle"

 style="display:block; text-align:center;"
 data-ad-layout="in-article"
 data-ad-format="fluid"
 data-ad-client="ca-pub-5653382914441020"
 data-ad-slot="7925631830"></ins>

<script>

 (adsbygoogle = window.adsbygoogle || []).push({});

</script>

usage

mlpack-config.cmake

#.rst:
# FindMLPACK
# -------------
#
# Find MLPACK
#
# Find the MLPACK C++ library
#
# Using MLPACK::
#
#   find_package(MLPACK REQUIRED)
#   include_directories(${MLPACK_INCLUDE_DIRS})
#   add_executable(foo foo.cc)
#   target_link_libraries(foo ${MLPACK_LIBRARIES})
#
# This module sets the following variables::
#
#   MLPACK_FOUND - set to true if the library is found
#   MLPACK_INCLUDE_DIRS - list of required include directories
#   MLPACK_LIBRARIES - list of libraries to be linked
#   MLPACK_VERSION_MAJOR - major version number
#   MLPACK_VERSION_MINOR - minor version number
#   MLPACK_VERSION_PATCH - patch version number
#   MLPACK_VERSION_STRING - version number as a string (ex: "1.0.4")


# UNIX paths are standard, no need to specify them.
find_library(MLPACK_LIBRARY
    NAMES mlpack
    PATHS "$ENV{ProgramFiles}/mlpack/lib"  "$ENV{ProgramFiles}/mlpack/lib64" "$ENV{ProgramFiles}/mlpack"
)
find_path(MLPACK_INCLUDE_DIR
    NAMES mlpack/core.hpp mlpack/prereqs.hpp
    PATHS "$ENV{ProgramFiles}/mlpack"
)


if(MLPACK_INCLUDE_DIR)
    # Read and parse mlpack version header file for version number
    file(STRINGS "${MLPACK_INCLUDE_DIR}/mlpack/core/util/version.hpp" _mlpack_HEADER_CONTENTS REGEX "#define MLPACK_VERSION_[A-Z]+ ")
    string(REGEX REPLACE ".*#define MLPACK_VERSION_MAJOR ([0-9]+).*" "\\1" MLPACK_VERSION_MAJOR "${_mlpack_HEADER_CONTENTS}")
    string(REGEX REPLACE ".*#define MLPACK_VERSION_MINOR ([0-9]+).*" "\\1" MLPACK_VERSION_MINOR "${_mlpack_HEADER_CONTENTS}")
    string(REGEX REPLACE ".*#define MLPACK_VERSION_PATCH ([0-9]+).*" "\\1" MLPACK_VERSION_PATCH "${_mlpack_HEADER_CONTENTS}")
    
    unset(_mlpack_HEADER_CONTENTS)
    
    set(MLPACK_VERSION_STRING "${MLPACK_VERSION_MAJOR}.${MLPACK_VERSION_MINOR}.${MLPACK_VERSION_PATCH}")
endif()

find_package_handle_standard_args(MLPACK
    REQUIRED_VARS MLPACK_LIBRARY MLPACK_INCLUDE_DIR
    VERSION_VAR MLPACK_VERSION_STRING
)

if(MLPACK_FOUND)
    set(MLPACK_INCLUDE_DIRS ${MLPACK_INCLUDE_DIR})
    set(MLPACK_LIBRARIES ${MLPACK_LIBRARY})
endif()

# Hide internal variables
mark_as_advanced(
    MLPACK_INCLUDE_DIR
    MLPACK_LIBRARY
)

From here

CMakeLists.txt

find_package(MLPACK REQUIRED)
MESSAGE( [Main] " MLPACK_INCLUDE_DIRS = ${MLPACK_INCLUDE_DIRS}") 
MESSAGE( [Main] " MLPACK_LIBRARIES = ${MLPACK_LIBRARIES}")  
# /usr/local/include
# /usr/local/lib/libmlpack.so

mlpack clustering

see mlpack clustering

kmeans

skip now.

meanshift

• mlpack_meanshift doc
• mlpack mean shift test

dbscan

• mlpack dbscan test

sklearn clustering

from sklearn.cluster import MeanShift
from sklearn.cluster import DBSCAN
from sklearn.cluster import KMeans

see sklearn clustering

opencv clustering

• cv::kmeans()

see opencv clustering

Reference

• mlpack
• mlpack.org
• mlpack findpackage
• mlpack mean shift
• dlib

History

• 20190520: created.

Copyright

• Post author: kezunlin
• Post link: https://kezunlin.me/post/1cd6a04d/
• Copyright Notice: All articles in this blog are licensed under CC BY-NC-SA 3.0 unless stating additionally.