A julia package containing the code to create the figures from On the adaptive spectral approximation of functions using redundant sets and frames by Vincent Coppé and Daan Huybrechs.
Run the notebooks to create the figures yourself, its prerequisit is a Julia 1.2 kernel. Two notebooks are presented that create similar figures. The notebook adaptivity figures, no AZ uses truncated singular value decompositions to solve all systems. This is computational heavy, but the most general approach. The notebook adaptivity figures uses the AZ algorithm to speed up the experiments. More detail on the AZ algorithm is provided in The AZ algorithm for least squares systems with a known incomplete generalized inverse. To appear in SIAM Journal on Matrix Analysis and Applications.