Avoid recalculating X^q when f(p): X.^1 ... X.^p

[taqtiqa-mark]

@stevengj

if you only get cache misses infrequently, then the performance of the ^n exponentiation is probably irrelevant, and I would tend to just use pc.x .^ n to maximize accuracy.

Thank you for the heads-up about performance and accuracy, for those following along, see @stevengj’s efficient X^p, p=1...n script here.

Just to be clear, the ‘cache’ misses you mention refer to the PowerCache in Max’s code above?