How sansk552/olist-ecommerce-sql-python-analysis works
This repository is an Olist e-commerce analytics project, not a general application. Its documented workflow turns transactional data into SQL-based analytical datasets and business metrics, explores them in Python, and extends the analysis with bad-review prediction. The README also describes the prediction features, model comparison, and SHAP explanations. Source excerpts were unavailable, so relationships below follow only the README; individual SQL-query responsibilities and notebook internals are not inferred.
Main entry points: olist_analytical_dataset.csv Analytical datasets · master_dataset.sql Dataset construction · sql1.sql SQL metrics · Olist_analysis.ipynb Exploratory analysis
Components 10
Data preparation
- Dataset construction
master_dataset.sql - Analytical datasets
olist_analytical_dataset.csv - Complete dataset
olist_complete_analytical_dataset.csv
Business analysis
- SQL metrics
sql1.sql - SQL result files
sql1_output.csv - Exploratory analysis
Olist_analysis.ipynb - Business insights
Review prediction
- Bad-review prediction
olist_review_prediction.ipynb - Model explanations
olist_review_prediction.ipynb
Other components
- Olist data
Connections 8
- Olist data → Dataset construction: transforms
README.md - Dataset construction → Analytical datasets: produces
README.md - Analytical datasets → Exploratory analysis: analyzes
README.md - Analytical datasets → SQL metrics: queries
README.md - SQL metrics → SQL result files: exports metrics
- Exploratory analysis → Business insights: derives
README.md - Complete dataset → Bad-review prediction: supports
- Bad-review prediction → Model explanations: explains predictions
README.md