Hello guys, what is the fastest way to sort an Int...

# announcements

[g4sarma]

Hello guys, what is the fastest way to sort an Int[] ?

[diesieben07]

Just call sort 😛

[g4sarma]

I am aware about the sort() in kotlin. is that good enough to take care of the performance @evanchooly

[diesieben07]

Yes.

[diesieben07]

https://docs.oracle.com/javase/10/docs/api/java/util/Arrays.html#sort(int%5B%5D)

[diesieben07]

You can usually trust the JDK to know what they are doing 🙂

[g4sarma]

Cool, thanks for that link to the doc

[evanchooly]

“do the simplest thing that can possibly work.” then optimize as necessary.

[Shawn]

you can always just ctrl/cmd+b into the definition of the method in question…

[Shawn]

many of the standard extensions provided on collection types are backed by stuff in the java stdlib anyhow

[Shawn]

and the docs will be right there

[Jeremy Rempel]

if its a known range ie, 1-100 you can do in o(n) time at cost of memory

[Jeremy Rempel]

but otherwise quicksort in nlogn

[g4sarma]

you mean the dataset @Jeremy Rempel?

[g4sarma]

or the size

[Jeremy Rempel]

the values, say all values in int[] were between 0-100

[Jeremy Rempel]

and you had a million of them

[g4sarma]

Okay, sp it i had a larger range of values say -1000 to 1000

[g4sarma]

quicksort it is ?

[Jeremy Rempel]

no, 0-100 is just a number. its just a defined range

[g4sarma]

That's what i meant

[Jeremy Rempel]

you'd loop through your array once and use a map to store map[i] = cnt

[Jeremy Rempel]

in O(n) time

[Jeremy Rempel]

then map[i] will represent the number of elements at i

[Jeremy Rempel]

then second loop from 0..range and copy values into a new array

[Jeremy Rempel]

O(n) time

[Jeremy Rempel]

you are creating a map of size n so its a tradeoff between time and memory