Cocktail Clustering Project

Overview

It's a recruitment project for the Solvro Scientific Club at Wrocław University of Science and Technology. It focuses on exploratory data analysis (EDA) and clustering of a cocktail dataset. The task involves conducting EDA, cleaning and preprocessing the data and performing clustering using KMeans and DBSCAN to identify unique clusters of cocktails based on their ingredients and other characteristics.

Data 📋

The dataset contains cocktails with associated ingredient lists and characteristics. It's stored in a .json file located in the data folder. This is loaded and processed into a DataFrame format for easy manipulation and analysis using Pandas.

Repository Structure

data/: Raw and processed datasets.
scripts/: Scripts for data preprocessing and clustering.
notebooks/: Jupyter notebooks for EDA, cleaned data visualization and clustering visualization.
requirements.txt: Dependency management files.

Quick start ⚡

To replicate the experiments, follow steps listed below (in cmd).
As an alternative you can do it using conda.

1. Clone the Repository

git clone https://github.com/DominikaStefaniak/ml-solvro-recuitemnt.git
cd ml-solvro-recruitment

2. Set up the environment and install libraries

If you want to do it in cmd:
First create venv:

python -m venv venv

Activate venv:
For windows:

venv\Scripts\activate

For macOS and Linux:

source venv/bin/activate

Install the dependecies from requirements.txt
note: the library versions are compatible with python 3.12

pip install -r requirements.txt

Possible issues:

• in venv "bin" file created instead of "Scripts" for windows: use command prompt or power shell, if you are then python may have been installed incorectly
• error while libraries instalation: make sure to use python 3.12 or check the libraries versions compatible for your python version and change them in requirements file

3. Steps of the experiments

Remember to always delete corresponding data file before running any of the .py files

• data analysis in eda.ipynb file
• data cleaning and preprocessing in data_preprocessing.py
• data visualization in data_visualization.ipynb
• additional preprocesing and clustering models with quantitative evaluation in clustering.py
• qualtitative evaluations with summary in clustering_results.ipynb