Quick coding experiments and small projects with some real outputs.

You can open the repository in server-hosted online environment, Binder from here:

1. alcoholism

   • Promoting dry January, we developed full pipeline of basic data-science model predicting alcohol consumption from publically available data.
   • First get the data from World bank using python API, wbgapi. Second clean and do preliminary analysis using pandas, together with visualization in seaborn. Third, develop linear regression model in scikit-learn, explore spurious correlation, overfitting, etc.

2. streamlit-dashboard

   • Minimalistic way how to create streamlit interactive dashboard using examplery ML Iris dataset.
   • You see that in 130 lines of code, you train model, visualize results, and create interactive environment for the user to look into your KMeans clustering.

3. r2py

   • In 2022, introductory course to R has been given by @ramiromagno and @iduarte at CCMAR using cute synthetic crab dataset. All resources in R can be found here.
   • This project translates and compares how the same things are done using python, specifically pandas.