Tools for data science
A collection of tools and utilities I have written that I find useful for data science
Jupyter Notebook with examples
Tools
- Random Forest model interpretation using the awesome treeinterpreter package
- Gives model decision process for individual examples
- Automating the PowerPoint process using python-pptx
- Write Python code to save model result to PowerPoint
- Plotting functions
- ROC
- Random Forest Feature Importances (plus error bars)
- Functions for writing SQL code using Python
- Data imputation
- Dummy variables (use N most frequent column values for dummy cols)
- Date features (day of week, month, day of month)
- Agg features min,max,mean,median
- Data prep for modelling using Pandas