Can chemometrics be used to guide the selection of suitable DFT functionals?

•Selection of suitable DFT functionals can often be a time consuming process.•The classical approach of testing each molecule with multiple functionals is not practical for high throughput screening.•QSPR models are used to rapidly evaluate which functionals from an available set will be more suitable for a given problem.•The selected application is the prediction of the maximum absorption wavelength for azobenzene and anthraquinone dyes.

This article investigates whether it is possible to use chemometric methods, in particular quantitative structure–property relationships (QSPRs), to decide which DFT functional among the available choices, is best suited for a given problem. Here, the focus is on identifying the functional capable of obtaining the best theoretical approximation to experimentally observed maximum absorption wavelength λmax for a series of azobenzene and anthraquinone dyes. The results suggest that a QSPR-based approach can be a valuable strategy for avoiding unnecessary DFT computations, thereby facilitating rapid screening of structures.