An implementation of fBLS, a fast-folding-based BLS periodogram, described in ArXiv: 2204.02398.
The core of fBLS consists of a dynamic programming algorithm, akin to the Fast Folding Analysis, that was proposed for pulsar search. This algorithm simultaneously generates folded profiles for a dense grid of trial periods in linear time. The grid of trial periods is analyzed in a search for box-shaped signals (BLS), thus producing a BLS-periodogram.
Use Conda to set up the required environment from the yml file. The fBLS_illustration notebook aims to describe the algorithm. fBLS_example executes the code for a real example from Kepler.