Quicksilver #2: Sequence (Results)

[Vexorian]

Rules

• One week to solve.
• Submit solutions to the pastebin, make sure to have a [quicksilver] prefix.
• I will test each submission against a big set of test cases.
• 3 fastest solutions (those which take the least time) win. (It is not a solution if it doesn't show the correct result for any test case)
  • 1st place: 20 rep
  • 2nd place: 15 rep
  • 3rd place: 10 rep
• If you use global variables in your submission include them in a globals block.
• Your submission should not get into the op limit in any case, in order to override the limit, you cannot use TriggerSleepAction nor PolledWait, you have to use ExecuteFunc/TriggerExecute. It CAN get into the op limit if you call the function twice, but an instance of the function should never get into the limit.

Problem #2: Sequence
I have a global integer array udg_input and udg_count , udg_input contains udg_count elements, if udg_count is 7 then udg_input has the indexes 0..6 assigned.

udg_count is a number between 2 and 200
udg_input contains postive integer numbers not greater than 10000000 . No element appears in the list more than once. (no repetitions)

The objective is to make a function:
function Consecutive takes nothing returns boolean
It returns true if and only if the elements of the array can form a sequence of consecutive numbers.

Examples

To make the examples, I will use [i0 i1 i2 ...] notation to abbreviate the array.

Then Consecutive([4 5 6]) means:

JASS:

    set udg_count=3
    set udg_input[0]=4
    set udg_input[1]=5
    set udg_input[2]=6
    set b=Consecutive()

All right:

Consecutive( [1 2 3 4 5 6] ) : returns true
Consecutive( [1 2 3 4 5 7] ) : returns false

Consecutive( [6 2 5 4 3 1] ) : returns true
Consecutive( [3 2 4 5 6 1] ) : returns true

Consecutive( [3 2 4 5 6 1] ) : returns true

Consecutive( [102 103 104 105 106 107 108 109 110 111 112 113] ) : returns true


Consecutive( [100013 100002 100003 100004 100005 100006 100007 100008 100009 100010 100011 100012] ) : returns true

Consecutive( [100013 100002 12 100004 100005 100006 100007 100008 100009 100010 100011 100012] ) : returns false





The results

• 1st: CaptainGriffen : 0.032236452s
• 2nd: SeasonsOfLove : 0.048517480s
• 2nd: Blu_da_noob : 0.048734676s

(the difference was way too small, so I decided to declare a tie, (each gets 17 rep)

Other submissions:

• Anitarf: 0.053227096s
• Daelin: 0.053475628
• Acehart : 0.054932712s
• emjrl3 : 2.485663648s

disqualified:

• INfraNe : Wrong output
• BertTheJasser: Op Limit exceeded
• ChuckleBrother: Op Limit exceeded

The solution
It was really easy, there are no repetitions, then you can just get min and max values of the array and check if their substraction matches (count-1)

There was an alternative solution in which you only get the minimum and the sum then check with Gauss' formula for sum of consecutive numbers.

Both solutions ended up being as efficient, but the min/max version has an advantage and it is that with little changes you can make it stop the iteration when you can be sure it is wrong.

Otherwise, when 2 persons submitted the same algorithm, little details in the implementation decided who wins.

emjrl3 is the only person who used a different, heavier algorithm that didn't get to the op limit/had issues.

Extra Kudos go to anitarf for being the first person to submit a O(n) algorithm, the rest eventually updated theirs later.

So here is a solution:

JASS:

function Consecutive takes nothing returns boolean
 local integer c = udg_count
 local integer i = c-2
 local integer t
 local integer l = udg_input[c-1]
 local integer m = l

    loop
        set t = udg_input[i]
        if l > t then
            if t + c > m then
                set l = t
            else
                return false
            endif
        elseif m < t then
            if l + c > t then
                set m = t
            else
                return false
            endif
        endif
        set i = i - 1
        exitwhen i == -1
    endloop

    return true
endfunction

[Chuckle_Brother]

Heh, looks like fun.

Are use lazy types who didn't bother with contest #1 allowed in?

[Vexorian]

Each quicksilver round is independent of the others.

[The)TideHunter(]

Originally Posted by Vexorian

Consecutive( [3 2 4 5 6 1] ) : returns true

Thats a sequence?
How'd you get that?

[Chuckle_Brother]

Rearrange it: 1,2,3,4,5,6

Edit: Good, this one looks like fun(by which I mean easy:)) so I think I will do it.

[BertTheJasser]

No big deal, took me some minutes.

[Chuckle_Brother]

Yeah, same. I think Vex is losing it. It being his edge.

Edit: Where be your entry? Even if this is an easy problem you may as well scoop up some rep while you can.

[BertTheJasser]

Originally Posted by Chuckle_Brother

Edit: Where be your entry? Even if this is an easy problem you may as well scoop up some rep while you can.

What?

[blu_da_noob]

Note: The challenge may not be simply solving the problem, but doing so in the fastest possible time. Algorithm efficiency is important.

Chuckle: Check your link, it has a message. The whole point isn't to post your solution for everyone to see, which was an absolutely marvelous idea there, because it's a competition. Everyone is supposed to submit a solution alone, without any help from other people so that it can be judged as a representation of your abilities.

[Vexorian]

yes?

[blu_da_noob]

We keep doing stuff at the same time which confuses the others' action >.<

So thanks for the 'apparently' psychic answer :P

(Question was: can we modify the data in udg_input in our function)

[Captain Griffen]

Submitted.

Handy Tip Bubble sort is by far the most effective for this.

[Anitarf]

Submitted.

I had to edit it just a few minutes afterwards, serves me right for not testing it. I hope Vex has the latest version.

Anyway, this one was almost absurdly simple.

[Chuckle_Brother]

I thought we were supposed to put it in the pastebin, no?

because it's a competition

Oh well, it was simple so unless someone is REALLY bad, they would know how to do this.

[Vexorian]

Yes, you were supposed to put it in the pastebin, but not link it.