This project is a personal journey in learning how to read and work with sparse matrices in the NPZ format using Golang. It includes a simple Gin API that scores Zhang's Metric based on a list input in JSON format.

• Go 1.21

1. Clone the repo

   git clone https://github.com/HeyLucasLeao/zhangmetric

2. Install Go packages

   go get .

1. Setting Dataset

   The Kaggle groceries dataset is a viable option for workaround in this project. However, any historical data that can be transformed into a CSR Matrix can also be utilized. For more information, see the eda_example.ipynb.

   Groceries Dataset: https://www.kaggle.com/datasets/irfanasrullah/groceries

2. Setting .env

   After having the CSR Matrix witch respective ordered column products, insert the filenames inside a .env

3. Request on API

   With running the project, send a GET request to http://localhost:8080/score, with a JSON body having the names of product you want to score. Example:

   [
   "whole milk",
   "cereals",
   "tea",
   "yogurt",
   "waffles"
   ]
   

Lucas Leão - [email protected]