Help using GMP in C: How to take non-integer power?

[launchcode]

Hi,

Can anyone tell me how to take a non-integer power using GMP in C?

For example:

y = x ^ 0.2,

where x is a GMP float.

Thanks.

[Krupski]

Hi,

Can anyone tell me how to take a non-integer power using GMP in C?

For example:

y = x ^ 0.2,

where x is a GMP float.

Thanks.

Why do you need so much precision? Isn't an ordinary float good enough?

-- Roger

[johnl]

Edit:

Whoops, wrong thread. Sorry.

[PythonPower]

Well one method I can see involves converting an exponent in the form a^b into a function with exp(). Simply put, a^b = exp(b*ln(a)).

But that relies on the functions exp() and ln() which don't exist. You can implement these yourself in various ways and the computational complexity will be the worst complexity of either function.

Here are some references for starters:
http://en.wikipedia.org/wiki/Exponential_function#Continued_fractions_for_ex
http://en.wikipedia.org/wiki/Logarithm#More_efficient_series

Hopefully there is a less convoluted way but in absence of that, you have my thoughts. I'll check some other ideas out.

[casevh]

If you need to do it in C, I would use mpfr. It uses GMP for the underlying calculations but provides a much broader set of functions.

You can also use mpmath. It is written in Python but can use GMP for some operations.

casevh

[PandaGoat]

I believe the previous poster hinted to what I will suggest. Search the following tome for what you seek: http://gmplib.org/manual/Float-Arithmetic.html#Float-Arithmetic. I even turned the pages for you.

[launchcode]

Thanks guys. Much appreciated.