Skip to content

hkumar747/blocpower-buildings

Repository files navigation

BlocPower's building-level energy data

This repository hosts Jupyter notebooks that use BlocPower's massive building level energy data for 120+ million buildings, hosted on EIDC Redivis. They demonstrate how to perform several common operations, including data loading, aggregation to state/zipcode levels and visualization. They also discuss more advanced decision support frameworks, such as mathematical optimization for recommending energy policy interventions.

Notebook 1 loads and summarizes data by state, comparing the speed of Pandas, Polars and PySpark.

Notebook 2 shifts the focus towards optimization techniques, utilizing Gurobi to apply mixed integer programming (MIP) to select the most promising buildings for retrofitting given a fixed budget, for example, installing heat pumps.

image

Code for the BlocPower User Guide, which compares the building data in this dataset to other ground-truth datasets for building counts and energy use.

Getting Started

To explore the methodologies or replicate the analysis, just download and run the notebooks. It requires you to have API keys and authentication for the following:

  • Gurobi (free for students)

  • Redivis

About

Repository for code to process BlocPower building data.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •