HydroBlocks relies on a number python libraries. To make this straightforward use conda (http://conda.pydata.org/miniconda.html) to install all the packages. Here are the steps to install the model.
- conda install gdal
- conda install scikit-learn
- conda install netCDF4
- git clone https://github.com/chaneyn/HydroBlocks.git
- cd HydroBlocks
- python setup.py
- cd ..
- git clone https://github.com/chaneyn/geospatialtools.git
- cd geospatialtools
- python setup.py install
- cd ..
Note: If you have the intel mkl library on your machine, we recommend that you set the mkl_flag variable in setup.py to True.
To run the model on a test dataset, do the following:
- wget https://www.dropbox.com/s/tw4z4rf9ol6p24x/HB_sample.tar.gz?dl=0
- tar -xvzf HB_sample.tar.gz
- cd HB_sample
- python ../HydroBlocks/Preprocessing/Driver.py metadata.json
- python ../HydroBlocks/HydroBlocks/Driver.py metadata.json