Energy Consumption Impact Analysis (SQL + Python Pipeline)

This repository contains an end-to-end data pipeline + analysis project:

• Python ETL to build a normalized SQLite database from raw municipal/utility/property files
• SQL transformations to create analysis-ready tables/views
• Python analysis + visualization to compare energy consumption patterns across groups

A full write-up is included in the report:

• docs/01 Shade, Savings, and Sustainability Zumbulev.pdf

---

What this project does

Goal: Build a clean relational dataset and analyze how electricity consumption changes before vs. after events, comparing:

• Tree removal permits
• Energy Use Data

The pipeline includes **private addresses, which is why the raw data and derived database are treated as sensitive.

---

Repository structure

• src/ – Python scripts (ETL + analysis)
  • build_database.py – builds the SQLite database from raw files
  • analysis.py – runs analysis and generates a plot
• sql/ – SQL logic
  • baseline_analysis.sql – builds analysis-ready relations
• docs/ – report PDF
• data/ – local raw data (NOT included in this repo)

---

Data privacy (why data/ is not included)

This project uses raw data that includes addresses and other sensitive fields.
Therefore, raw files and the generated SQLite database are intentionally excluded from GitHub.

You must place raw data locally under data/ to run the pipeline.

Expected files in data/ (based on the scripts):

• CitizenserveReport.xlsx
• GRU_Customer_Electric_Consumption_2012-2022.tsv
• Property.txt
• ImprvAttributes.txt
• ImprvDetails.txt

Also note:

• The ETL may generate address_cache.csv (contains address strings) — it is ignored by design.

---

How to run

1) Create environment + install dependencies

pip install pandas numpy matplotlib seaborn rapidfuzz openpyxl