How can you calculate percentiles from a cdf?

Jobocop · #1 $Old$ October 28th, 2009, 06:52 PM

I'm having some trouble understanding how to calculate the percentile from a cdf, i realize that a percentile ( ie. percentile .95) is the value in which 95% of all the probabilities contained within the function fall below - but how do you calculate that value from a function?

the fuction i was given is : ky(1-y) where 0>y>1 , btw this is my first post so i didnt know how to do a greater than or equal sign.. but that's what it's supposed to be! thanks

And if Mr.Fantastic is reading this, you've already saved my life a couple times. So thank you

mr fantastic · #2 $Old$ October 28th, 2009, 10:26 PM

Quote:

Originally Posted by Jobocop $View Post$

I'm having some trouble understanding how to calculate the percentile from a cdf, i realize that a percentile ( ie. percentile .95) is the value in which 95% of all the probabilities contained within the function fall below - but how do you calculate that value from a function?

the fuction i was given is : ky(1-y) where 0>y>1 , btw this is my first post so i didnt know how to do a greater than or equal sign.. but that's what it's supposed to be! thanks

And if Mr.Fantastic is reading this, you've already saved my life a couple times. So thank you

I save many lives. Every day.

Is $f(y) = ky(1-y)$ where $0<y<1$ the pdf?

If so, then first you must get the value of k by solving $\int_0^1 k y(1 - y) \, dy = 1$ .

Then the 95th percentile is the value of $a$ such that $\int_0^a k y(1 - y) \, dy = 0.95$ .