Skip to content
Andrei Novikov edited this page Mar 24, 2017 · 12 revisions

Supported Operating Systems

Prerequisites

  • Python 3.4 or greater (64-bit version);
  • C/C++ compiler that support (C++ 14);
  • Python modules: scipy, matplotlib, numpy, PIL;

64-bit issue

64-bit version is required for core library ccore, but it is possible to run vast of majority algorithms without the core using special flag in algorithm or model constructor, for example:

# Read sample for file 'path_sample'
sample = read_sample(SIMPLE_SAMPLES.SAMPLE_SIMPLE2);

# Create instance of OPTICS algorithm for cluster analysis
# To avoid usage of core library set 'ccore' to False state  
optics_instance = optics(sample, 3.0, 3, ccore = False);

Be aware that in this case execution time will increase, because only python implementation is used. To achieve maximum performance 'ccore' library should be used.

Be aware that basically 'ccore' is False by default for vast of majority algorithms and models in pyclustering.

Clone this wiki locally