This HPAT tutorial makes use of hpat (latest), numpy, pandas, daal4py (latest) and their dependences.

The easiest and most reliable way is to create a python environment using conda:

• Linux: conda create -n hpattut -c ehsantn -c numba/label/dev -c defaults -c intel -c conda-forge hpat daal4py pandas=0.23 blas=*=mkl jupyter notebook
• Windows: conda create -n hpattut -c ehsantn -c numba/label/dev -c defaults -c intel hpat daal4py pandas=0.23 blas=*=mkl jupyter notebook

Then activate the environment

conda activate hpattut

and you are ready to start the tutorial!

The main material is provided as juypter notebooks. To get started simply type

jupyter notebook

The main HPAT tutorial is in the notebook hpat.ipynb. An example of an advanced analytics code is provided in intraday_mean.py. A more complete notebook about using daal4py with HPAT can be found in daal4py_data_science.ipynb.