In the course I assist this spring (Simulation Methods in Finance and Insurance), there are plenty of places where one has to inverse a function.

For instance inversion method of generating pseudo random numbers from a general distribution, which requires to use an inverse c.d.f. (or quantile function). Further on, one has compute a so-called generator of Archimedean copula. Even for calculating a value at risk (VaR) the inversion might be useful. Of course, one can argue that all these methods have already been implemented in R, but the story is rather about inversion. In R there are no built-in function operator that returns a inverse of mathematical functions. This post propose a rather elegant solution (with corrected syntax):

inverse <- function(f, lower, upper) { function(y) uniroot(function(x) f(x) - y, lower = lower, upper = upper)[["root"]] }

Instead of explicitly define an inverse function, now all mathematical functions can be inversed by a line of code.

log2 <- inverse(exp, 0.1, 10) log2(exp(1))

Please, note that I avoid default values for arguments `lower`

and `upper`

, ’cause they really require a care.