Gaia Flux Variability — InterSystems Employee Programming Challenge #1

A Python solution that identifies highly variable astronomical sources from Gaia DR3 epoch photometry data.

Submitted for the InterSystems Employee Programming Challenge #1.

What It Does

The program processes 20 Gaia DR3 epoch photometry files and produces output.csv listing every source whose BP or RP band flux varied by more than 100% between its minimum and maximum valid observations.

For each source:

1. Extracts the bp_flux and rp_flux arrays (per-observation brightness measurements in Gaia’s blue and red photometer bands).
2. Filters out invalid values — missing, null, NaN, and non-positive fluxes.
3. Computes the percentage change for each band: ((max − min) / min) × 100.
4. Takes the larger of the two percentages as the source’s variability score.
5. Emits the source to output.csv if the score exceeds 100%.

57,099 sources qualify across the 20 input files.

Prerequisites

• Git
• Docker Desktop

Installation

Clone the repository:

git clone https://github.com/munraj-bal/intersystems-challenge-1-mbal.git
cd intersystems-challenge-1-mbal