Formally, the integral of a function *f* is a function denoted by
*f* which satisfies
*d* (*f* )/*dx* = *f*; note that such a function
is only determined upto a constant since a constant has derivative 0.

Similarly, if *f* (*x*) is a function, its formal inverse is a function
*g*(*y*) so that
*g*`o`*f* = identity. Finally, if *f* (*x*, *y*) is a
function of two variables, we can look for a function *g*(*x*) so that
*f* (*x*, *g*(*x*)) = 0 identically.

Algebraically this is all we need. We can easily compute the values of derivatives at various orders of the new functions in terms of those of the old functions.

Analytically, we need to show that such functions exists under certain reasonable conditions on the given data. For exmaple, we have already constructed an integral for a polynomial function.

Here is an example of inversion: