fastcluster: Fast hierarchical clustering routines for R and Python

Copyright:
  * Until package version 1.1.23: © 2011 Daniel Müllner <https://danifold.net>
  * All changes from version 1.1.24 on: © Google Inc. <https://www.google.com>


Installation
‾‾‾‾‾‾‾‾‾‾‾‾
Installation procedures were tested under 64-bit Ubuntu. CRAN also hosts
precompiled binaries (of the R library, not the Python module) for Windows and
OS X. In principle, it should be possible to install the fastcluster package
on any system that has a C++ compiler and R respectively Python with NumPy.
There are no unusual libraries needed to compile the package, only the STL
library, which every C++ compiler should have by default.

Please send me feedback if you accomplish to install the fastcluster package on
a certain platform but needed to tweak the configuration! I will update the
installation instructions and modify the package if needed (eg. include the
right compiler flags for various operating systems).


Installation for R
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
Enter the command install.packages("fastcluster") in R, and R will download the
package automatically, then install it. That's it!

If this does not work, please consult R's help function by typing ?INSTALL from
within R or read the “R installation and administration” manual:

    http://cran.r-project.org/doc/manuals/R-admin.html#Installing-packages

For manual download, you can get the fastcluster package from the download page
at CRAN:

    http://cran.r-project.org/web/packages/fastcluster/

You may need to start R with administrator rights to be able to install
packages. There are ways to install R packages without administrator privileges
in your user directories. See this help page for example:

  http://csg.sph.umich.edu/docs/R/localpackages.html


Installation for Python
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
Make sure that you have both Python and NumPy installed.


1. On all platforms
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
If pip is installed, type

    pip install --upgrade --user fastcluster

in a terminal, which automatically downloads the latest version from PyPI,
compiles the C++ library and installs the package for a single user without
administrator rights.

If this works, there is no need to follow the alternative steps below.


2. Microsoft Windows
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
Installation files for Windows are stored on PyPI:

    https://pypi.python.org/pypi/fastcluster

Christoph Gohlke also provides installation files for Windows on his web page:

    http://www.lfd.uci.edu/~gohlke/pythonlibs/#fastcluster


3. With setuptools
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
If pip is not available but setuptools, type

    easy_install --upgrade --user fastcluster

in a terminal.


4. From the source package
‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
If you have not done so already, download the fastcluster package from PyPI
here:

    http://pypi.python.org/pypi/fastcluster/

Open a terminal, go to the directory with the downloaded file and extract the
contents of the archive with:

    tar -xvf fastcluster-(version).tar.gz

Alternatively, use your favorite archive manager for unpacking, eg. on
Windows. This will generate a new directory “fastcluster-(version)”. Switch to
this subdirectory:

    cd fastcluster-(...)

The source distribution on CRAN also contains the complete source files. See
the directory src/python there.

Now compile and install the Python module by:

    python setup.py install

You may need to precede this command with sudo or install the package in your
home directory, like this:

    python setup.py install --user

See the chapter “Installing Python modules” in the Python documentation for
further help:

    http://docs.python.org/install/index.html