Want to Try Theoretical Physics This Evening? I Built a Python Framework for That
Theoretical physics has a brutal cost of entry: years of prerequisites before you get to touch anything real. I wanted to lower that first step — not the science, just the step. So I built SFT (Spectral Flow Transform), a Python framework that treats complex systems through their spectral fingerprints: frequencies in audio, energy levels in physical systems, modes in simulations.
Three things it does well: predict how a spectrum shifts when parameters change, solve the inverse problem — find parameters that produce a spectral shape you want — and measure how complex a system actually is via spectral rank.
To prove it’s not a toy, I pointed it at SPARC, a public database of galaxy rotation curves. Across 107 galaxies, the normalized curves turned out to have low intrinsic dimensionality — and the second spectral mode correlated with surface brightness and morphology without ever being told about them.