A group of physicists and computer scientists has developed a machine learning strategy that can extract charge density wave (CDW) – an ordered modulation of electrons – and intra-unit-cell (IUC) parameters from high volumes of X-ray diffraction data at multiple temperatures. The team's approach, called X-TEC (X-ray diffraction temperature clustering), exploits the fundamental role that temperature plays in both long-range and short-range structural correlations.