#HTM Filetype

A corruption tolerant filetype for saving neurogenesis/neurodegeneration enabled HTM neural networks

##Prerequisites

###For Whole Project

• rtree:
  • Get from here

Rtree sets up a tree of rectangles to sort all the n-dimensional points into. This makes it so individual points can be found in O(log_m(n)) time.

###For Saving

• flatbuffers
  • pip2 install flatbuffers

Allows saving individual neural columns to files. (Not implemented yet.)

###For Testing and Visualization

• Anaconda Python 2.7
  • Get 2.7 version from here
  • Set up in Pycharm via File->Settings->Project:[project name]->Project Interpreter
• VTK
  • conda install -c anaconda vtk=6.3.0

Anaconda is useful for installing VTK. You could build VTK for python, but that's harder to do on Windows.

##Testing ###Point field testing VTK required.

All points should be arranged in rows within prism, save for a few outliers, and colors should show linear order.

It's harder to see linear order when a lot of points are generated, but there should be several lines visible on some of the faces of the generated prism.