Addictive Math Puzzles

For discussion of non-Wonderland topics - please read rules!

Moderators: ~xpr'd~, tyteen4a03, Stinky, Emerald141, Qloof234, jdl

Post Reply
User avatar
VirtLands
Rainbow Master
Posts: 756
Joined: Thu Dec 29, 2005 1:49 am

Addictive Math Puzzles

Post by VirtLands » Tue Feb 07, 2012 2:25 am

Hi there,

This is a type of math puzzle that I dreamt up.

(Later on I discovered that it was already discovered by others on the web, but I don't think that they also thought of the 2-D version. )

It's called "Maximum non adjacent subsequence sum".

Given an array A[..], filled with numbers *(that I chose),
your mission is to find the maximum sum of a SUBSET of array A[..]
with the constraint that no 2 numbers are adjacent to each other.
That is, - the SUM that is MAXIMUM wins and is the correct answer.


PUZZLE 1: -- Array A[..]
Image

PUZZLE 2: -- Array B[..]
Image

For puzzle 2, do the same thing.
Both arrays have only 16 numbers each.

Hints:
In puzzle 1, you may add the first "5" and "9" together (=14),
but you may not add that "5" to "8" since "5" and "8" are adjacent.
You may skip and choose any numbers that you wish as long as
they are not neighbors. MAXIMUM SUM wins.

Let's see how many get the correct answers for both puzzles.

There is even a slim possibility that one array will have 2 equal answers,
= a tie, but it's not likely, the odds are against it.

**************************************************

Try to solve the 2-Dimensional version: Image

It is a 10x6 arrray, the same rules apply,
extended UP-DOWN as well as LEFT-RIGHT.
No adjacent numbers may be summed together,
that is:

A number may not be summed with the one to its adjacent LEFT or
adjacent RIGHT.
A number may not be summed with it's immediate UP or DOWN.
The MAXIMUM SUM wins.

Any numbers chosen will have an effect on the
availability of others.

For example, in the first line, if you chose "41" then you will
not be able to choose "85" in the lower line because they are adjacent.

PUZZLE 3: -- Array C[..][..]
Image

This is so hard that you may have to design a computer program to solve it.
User avatar
VirtLands
Rainbow Master
Posts: 756
Joined: Thu Dec 29, 2005 1:49 am

maximum non adjacent sums solution {Puzzles 1 & 2}

Post by VirtLands » Tue Feb 07, 2012 10:28 pm

Image O.K. then, it looks like no one was able to solve the 1st two puzzles,
( Puzzles 1 & 2 );
I created the following Blitz3D program to solve Puzzles 1 & 2:

Enjoy.

The heart of the recursive function is

Return max( MaxSum(N-1), A[N]+MaxSum(N-2) ),

where N is chosen to be the last index into the array A[].
In this case, the last index is A[15].

And the answers to the Puzzlers 1 & 2 are:

Max non-adj sum for {puzzle 1} = 48
Max non-adj sum for {puzzle 2} = 623

Of course, there are two obvious drawbacks with the current code :shock:

A: It does not list which actual numbers were summed up, and
B: It does not take into account that there may be two or
more different sequences that add up to the same sum (= a tie).

When I get back, I'll post an improved version which details all solutions.

I also have to figure out how to solve Puzzle 3, which is 2D.
:)

There were other algorithms to solve the same thing, but my version
is the only one that I understand.
http://stackoverflow.com/questions/3733 ... n-an-array
http://nthrgeek.wordpress.com/2009/10/0 ... bsequence/

Blitz3D Demo:
http://www.blitzbasic.com/Products/_index_.php
Last edited by VirtLands on Sat Feb 11, 2012 10:32 pm, edited 1 time in total.
Muzozavr
Rainbow Spirit Chaser
Posts: 5648
Joined: Wed Jan 11, 2006 2:55 pm

Post by Muzozavr » Tue Feb 07, 2012 11:26 pm

Got 1 in 48, didn't try 2 and 3 yet...
EDIT: Got 2 in 623.

Solution for puzzle 1, left-to-right:
5+9+3+8+5+7+2+9=48
Solution for puzzle 2, left-to-right:
74+99+13+88+145+114+90=623
Rest in peace, Kym. I hardly knew ya.
Rest in peace, Marinus. A bright star, you were ahead of me on my own tracks of thought. I miss you.
User avatar
VirtLands
Rainbow Master
Posts: 756
Joined: Thu Dec 29, 2005 1:49 am

Muzozavr solved linear non-adjacent sums puzzles

Post by VirtLands » Wed Feb 08, 2012 10:05 pm

Wow Muzozavr, you're a brain; those answers are correct
for Puzzles 1 & 2. :)

I'm still working on Puzzle 3 myself, and will try to make some code
to solve it. Will be back later.
Last edited by VirtLands on Mon Feb 13, 2012 11:55 pm, edited 1 time in total.
Muzozavr
Rainbow Spirit Chaser
Posts: 5648
Joined: Wed Jan 11, 2006 2:55 pm

Post by Muzozavr » Wed Feb 08, 2012 10:22 pm

I actually got the first one before you even posted the max answer... it's just that 48 seemed ridiculously low to be the max answer. But it IS the max answer, as the program proves... so it was easy to find.

And puzzle 2 is still way easier than what I have to do for my science paper that I'm writing for school. And I have to write it faster than Sonic the Hedgehog runs. I thought I'd be able to talk with the teachers and postpone the project... but no, it's impossible.
Rest in peace, Kym. I hardly knew ya.
Rest in peace, Marinus. A bright star, you were ahead of me on my own tracks of thought. I miss you.
User avatar
VirtLands
Rainbow Master
Posts: 756
Joined: Thu Dec 29, 2005 1:49 am

"Maximum non adjacent subsequence sum" code update

Post by VirtLands » Sun Feb 12, 2012 12:11 am

I cleared away this sixth post to save space for the next post.
To see the old post as it was, click here:

http://i37.photobucket.com/albums/e73/V ... dPost6.jpg
Post Reply