Setup

Download and Install Anaconda from https://www.anaconda.com/products/individual

After installing add the following to ~/.bash_profile in terminal

export PATH=$PATH:/Users/<NT_USERNAME>/opt/anaconda3/bin/conda

The PATH is the installation location on MAC OSX

Download and Install PyCharm from https://www.jetbrains.com/pycharm/download/#section=mac

Environment Creation

Step 1 : In PyCharm, create a new project using Conda environment as below-

Step 2: Clone this repository in the project terminal

git clone [email protected]:Monetization/pricingml.git

Step 3: Move into pricingml directory

cd pricingml

Step 4: To install all required packages and dependencies, create the Conda environment using

conda env create --name pricexml --file environment.yml

Step 5: Activate environment using

conda activate pricexml

To deactivate environment

conda deactivate

Authentication

In the pricingml/utils directory, find the file credentials.

Enter username in the first line (ex: kashishchanana)
Enter pin followed by yubikey hash in the second line (ex: 123456yubikeycode)

Arguments

--connect
Enter y to set up a new connection

--feecode
Enter Fee Code to be Analyzed

--siteid
Enter Listing Site ID

--multi
Enter y for multivariate forecasting

--hourly
Enter the number of hours you want to aggregate the data on ex: 12 for 12 hourly analysis

--WoW
Enter y for Week on Week Analysis

--substier
Enter y for Analysis by SUBS_TIER

--model_name
Specify the ML model you want to run, use Prophet-Multi

Execution

Run the main file in the project terminal as below-

python3 main.py --connect 'y' --feecode 6001 --siteid 0 --multi 'y' --WoW y --hourly 12 --model_name Prophet-Multi --substier y