(2^p)*x - where used?

[Weiss]

came across this function in a math unit. I am building a bit of numerical environment for my work, part of it is collecting useful functions. I understand (2^p) could be a number of data points for Fourier transform. Where would I need it multiplied by another number? In itself (2^p)*x is a very steep climb.

BTW Google has no clue ( I love this part)

[MathMan]

came across this function in a math unit. I am building a bit of numerical environment for my work, part of it is collecting useful functions. I understand (2^p) could be a number of data points for Fourier transform. Where would I need it multiplied by another number? In itself (2^p)*x is a very steep climb.

BTW Google has no clue ( I love this part)

The function ldexp is often used to bring a float to a certain range. The calculation of elementary functions (i.e. log, exp, ...) in math libraries nowadays generally uses Tschbycheff series. They guarantee a specific minimum error for a float in a certain range - so you may find ldexp used in the preliminary setup phase. However, many math libraries directly manipulate the binary exponent of an IEEE-754 float to the same effect.