README.md

Eaze Code Challenge

This repo is mainly for Order Volume Prediction (Q5) part and documentation + thought process for all the challenges.
📖 Details / documentation of each questions can be found on the wiki page.

Dependencies + Compiler Installation

For Windows10 instructions :

1. Install C++ Compiler : mingw-w64

conda install libpython m2w64-toolchain -c msys2

2. Install PyStan ( conda or pip )

conda install pystan -c conda-forge

pip install pystan

3. Install Prophet ( conda or pip )

conda install -c conda-forge fbprophet.

pip install fbprophet

For more detail of how to set up - See here.

Param

• data_file : String value. Default Input file name - "hourly_volume.csv" that consists of "order_hour" and "order_volume".
  
• days_forward : An integer indicates the number of days to forecast order volumes out on hourly granularity. Default to 1 day.

Execute

Change input_data file on line #49 NEW_DATA_FILE_NAME and/or days_forward on line #47 NEW_NUM_DAYS_FORWARD.

run "question_5.py"

Output

An excel file named hourly_forecast_volume.xlsx would be generated under data folder.

Licensing

MIT License | Copyright © 2018 Eaze