*** This repository is still under development ***

Repo for analysis on hyper spectral image data using machine learning. Includes simulated image generators with accompanying evaluation models, and ensembling methods. As well as, various image processing tools, dataset I/O methods for creating and accessing datasets, and multiple visualization methods.

These instructions are for installing, and running a simple simulation

So most prerequisities can be obtained by using a standard anaconda install of python >=3.6.0 except a package called rasterio that is being used for some I/O of the dataset currently being worked on.

The explicit requirements are:

   'scipy>=1.1.0',
   'pandas>=0.23.0',
   'scikit-learn>=0.19.1',
   'scikit-image>=0.14.0',
   'scikit-hyper>=0.0.2',
   'tensorflow>=1.8.0',
   'h5py>=2.8.0'

As you will see in setup.py of this repo. Was also recently experiencing issues while setting up this environment for someone else, specifically with the rasterio->gdal->libgdal dependency stack. This issue is currently being worked on.

One issue with installing some of the dependencies leads me to recommend that you first build a python 2.x version and pip install rasterio with this python. Then create a virtual environment for a python 3.6 or 3.7 version pyenv virtualenv <version> <venv-name>. You can then complete the step of install rasterio, gdal, etc by pyenv activate <venv-name> pip install rasterio.

git clone --single-branch --branch rebuild-kspec https://github.com/tensor-strings/hsi_atk
cd hsi_atk
python setup.py --install