HMMlib is a C++ library for constructing and analyzing general hidden Markov models. The library consists of a number of template classes and generic functions, parameterized with the precision of floating point types and different types of hardware acceleration.

If you wish to use HMMlib in a Python program, we provide a set of Python bindings to hmmlib - see the examples below.

HMMlib is published under the GNU Lesser General Public License.

See also: HMMlib: A C++ Library for General Hidden Markov Models Exploiting Modern CPUspdf_logo



Compile and install:

In order to compile and install HMMlib, CMake must be installed on your system, and CMake must be able find the Boost C++ libraries on your system.

To compile and install HMMlib, download  the source code, by clicking one of the links above, and run the following commands in a terminal:

$ tar -xvf HMMlib-1.0.2.tar.gz
$ cd HMMlib-1.0.2
$ cmake .
$ make
$ make test
$ sudo make install

To use HMMlib you may now include eg. “HMMlib/hmm.hpp” in your program. See the examples below.

If Boost.Python is found in your Boost installation, python bindings will be generated in the file



$ make doc

to generate documentation for each class or find it here:




Using python bindings