added factorial function to int and float #3504
Conversation
|
Actually, you can melt these commits together before you create pull request. |
| @@ -96,6 +96,10 @@ struct Float | |||
| def self.from_io(io : IO, format : IO::ByteFormat) | |||
| format.decode(self, io) | |||
| end | |||
|
|
|||
| def factorial | |||
| self*Math.gamma(self) | |||
dylandy
force-pushed
the
factorial
branch
2 times, most recently
from
0b1b1bd to
de36fe3
Compare
|
@david50407 @Sija I just rebased those commits and recovered the other files that I accidentally formatted. |
| @@ -553,6 +553,10 @@ struct Int | |||
| self | |||
| end | |||
| end | |||
|
|
|||
| def factorial | |||
| self == 0 ? 1.0 : (1..self).to_a.map { |x| x.to_f64 }.reduce { |x, y| x*y } | |||
There was a problem hiding this comment.
and spaces here { |x, y| x * y }
|
fixed the format issue in int.cr. |
|
Thank you! But... I never used factorial in my life, and never saw it used except for code samples and blog posts. What's a real use case for factorial other than that? Also, factorial gets big pretty quickly, which means that using limited integers like |
|
If someone needs factorial, they can decide for themselves what number type they want to work with and (very easily) implement factorial accordingly. I think that this should not be a method on the number data types. Might be OK for the There's probably a good reason that Ruby doesn't have factorial. |
| @@ -553,6 +553,10 @@ struct Int | |||
| self | |||
| end | |||
| end | |||
|
|
|||
| def factorial | |||
| self == 0 ? 1.0 : (1..self).to_a.map { |x| x.to_f64 }.reduce { |x, y| x * y } | |||
There was a problem hiding this comment.
Just for the sake of code, we can avoid the arrays by using an iterator, and pass 1.0 to reduce to avoid casting all values to Float64:
self == 0 ? 1.0 : 1.upto(self).reduce(1.0) { |a, e| a * e }Still, this is the basic algorithm, exact but slow, and limited to 170!.
|
I follow up on what @asterite said. I'm not sure this is any common to be included in the standard library. It may prove to be more useful in an |
|
Thanks for reply, I'm working on some machine learning projects which needs a lots of performance and I was really upset about how slow Python and R did so I decided to move on to Crystal. But some of the basic and most use functions have missed here, so I just give it a try though. |
improved factorial function with avoid changing it into array reformating int
|
Closing. There's consensus this doesn't need to be in the stdlib. |
This method can behave better than Range#product method which can get value up to 170, the same behaviour as R lang does.
Also I added float number factorial calculation using gamma function.