You are not logged in.
24050
2*4 + 0 + 5 + 0 =13
24024
23998
2+3+9-9+8=13
23972
2-3+9+7-2=13
23946
(2*3)+9+(4-6) = 13
Hi anonimnystefy
Congrats for the solution.
Now, you get to propose your own puzzle
You just have to think about some numbers in your head
and complete the answer format.
A cool thing is that you can draw attention
to a number (between 1 and 99) with some special properties
(anyone who solved the puzzle can continue the chain)
***** ANSWER FORMAT *****
Answer to previous:
My numbers are between 1 and 99.
Glued together, they give the sequence ... [10 digits or less]
Their sum/mean is ...
Constraints/Hints:
#1
#2
Hi everyone
I saw the "add x" threads and thought about the following
chain puzzle. Hope some of you will like it :-)
______________The puzzle________________
My numbers are between 1 and 99.
Glued together, they give the sequence S (S must have no more than 10 digits)
Their sum (or mean) is ...
There are often multiple solutions, so one may add additional
constraints/hints (one of my numbers is a prime/Fibonacci number, etc.)
_____________A Simple Example___________
My numbers are between 1 and 99.
Glued together, they give the sequence 15723
Their sum is 81
Solution 1 : 1 57 23
Solution 2 : 1 5 72 3
I could have added the following constraint/hint
- At least 2 of of my numbers are Fibonacci numbers
to make Solution 2 the only valid answer
_____________First Puzzle____________
My numbers are between 1 and 99.
Glued together, they give the sequence 23931897
Their sum is 195
Constraints/Hints:
#1 At least three of my numbers are prime numbers
#2 At least one of my numbers is a 2-digit Fibonacci number
***** ANSWER FORMAT *****
Answer to previous:
My numbers are between 1 and 99.
Glued together, they give the sequence ... [10 digits or less]
Their sum/mean is ...
Constraints/Hints:
#1
#2
23920
(2*3) + 9 - 2 + 0 = 13
a la bobbym
Sorry for the late reply.
Great work. If you feel like it, you may try to combine, for your own fun, the different problems (reverse, noise, 1-2 or 1-3 digits for a first split) but you already solved my problems (I will finish the rest later).
And take a look @ the game (Mean Sumurai @ Google Play). If you don't have an Android, the iOS version will be out this month. I want this "number splitting" thing to have some visibility.
It's been great interacting with you guys. Very gifted guys, that's what you are. Thanks again.
... ... this post is starting to sound like I'm going away. I'm still a member of this forum, darn it :-D
I'm sticking around :-). If you have any questions, follow-ups, suggestions, I'll be happy to hear them.
Best,
A quick post to present the (probably) last parts.
I'll come back to your previous answers later today or tomorrow
____Part VII____
I no longer think it is that hard, compared to the "noise" digits.
It now strikes me as relatively easy (at least the simplest versions)
Anyway, here it is
Given the sequence, there is now some order involved because we want
to tag the first split (it will be the X of previous parts). You can see it as if any
split now has a boolean/binary attribute indicating whether it is the first or not.
Only one of the split can have this boolean at true.
For instance, with a sequence of length 3 and positions 123
1 2 3 --> 1 or 2 or 3 could be the first split
1 23 --> 1 or 23 could be the first split
12 3 --> 12 or 3 could be the first split
For now, let's keep it that simple
(actually, if the first split has to be the X, there are some
obvious constraints on the lengths of the splits)
The first split can be of length
(a) 1 or 2
(b) 1 or 2 or 3 (all the other splits would still be made of 2 digits at most)
___Part VIII__
Example: 32017 --> The first split defines X=20,
the sequence becomes 3[20]17 --> the splits 3 and 17 complete the splitting.
Note that 31 cannot be a valid split: 20 is in the way
The first split defines the target sum
8995383
115161845
The first split defines the target mean
369792
195457
There are much more difficult instances in Mean Sumurai but
these will be fine for a text-version (plus, I have to hurry)
later
Terrific.
Congrats to both of you. The formula I had looked like the one from the post #27 but used intermediate variables for (1+sqrt(5))/2 and (1-sqrt(5))/2 so confirming post #27 could not be done by a quick look.
And yeah, anonimnystefy, your new formula does look much nicer than the previous one.
I don't know bobbym's education but both of you are obviously gifted people. There's still V-b to solve but truth is I'm mighty busy these times so if you have an answer, don't expect a quick confirmation. I also have one (may be 2) related problems. I will present it (them) straight away after an answer for V-b
One last point: I presented this number splitting formulation on the math section of Reddit (don't trust Reddit caricature in the media, many of the guys in that forum are maths Ph.D.s) and it may turn out that you guys, along with a few other people close to me, have been working on a new and possibly research-worthy problem. Now, if one of us could spot an interesting link with any sort of existing concept or phenomenon, the whole thing could become very exciting. Keep your eyes open.
Best,
anonimnystefy,
that was super-fast. Much faster than expected.
Truth is I can't even say right now whether your answer is correct or not.
Do you have a recursive form? It would be easier for me to check.
My gut feeling tells me that your formula is right.
I'm quite busy right now so give me 1-2 days to get back to you.
Sorry :-(
I have to ask: What is your educational background?
Congrats bobbym for the splittings. Only one left.
I think you'll enjoy the game. Go get it :-)
Thanks bobbym
It was actually the afternoon and I had only 2-3 hours of sleep the night before.
I love my little boy more than anything but he is not exactly helping with my on-off insomnia problems :-).
Back to business
New people on the thread, please go read the first posts. Otherwise, you may not get what's going on.
We've been discussing a new(?) kind of splitting and I've been trying shameless plugs of Mean Sumurai,
an android game based on it ;-D. bobbym and anonimnystefy have solved all my challenges up to now.
___________Part V: The formulas
Given a sequence S of digits, and splits of at most 2, what is the number of possible
combinations if we impose
(a) 1 digit to be left out of the splitting
(b) 2 digits to be left out of the splitting
Problem is that the position of the noise digits affects the number of possible splits
because a split is always made of consecutive digits.
Let me give you examples with a sequence of length 6
We have as positions 123456
You have to consider that, every position, if left out, defines a different space of combinations.
For instance, if left out, the digit at position 1 defines the sequence 23456 (5 digits)
which, as we saw in Part I, gives F6 combinations.
Whereas the digit at position 3 would define 2 distinct sequences 12 and 456
and thus F3 * F4 combinations
Good Luck! You will need it ;-)
[F for Fibonacci]
________Part VI: The games
Example:
120 as a sum for 834978 --> 8+34+78 = 120 (9 left out)
Challenges
110 as a sum for 872992 (Leave out 1 digit)
92 as a sum for 43971296 (Leave out 2 digits)
17 as a mean for 771133843 (Leave out 1 digit)
23 as a mean for 497353 (Leave out 2 digits)
52 as a mean for 5989652 (Leave out 2 digits)
In Mean Sumurai, the games are presented in a more gamer-friendly way,
along with (depending on the difficulty level) a calculator option that
displays the current computations. Text-version like the ones above are probably harder
to solve but they are doable.
Best,
****************************************************************************
[Note for bobbym and anonimnystefy: there are no right to left considerations in these parts]
****************************************************************************
oh sorry. probably a lack of sleep.
Great Job bobbym! I'm definitely impressed by this forum. You're serious about your fun :-)
Well, I'll get some rest, go out a bit and probably post Parts V and VI
by this evening or tomorrow morning.
Best,
Hi bobbym
You don't have it for now.
Your formula must allow computation of
the number of combinations for any n.
Cheers
Congrats anonimnystefy
You're really skilled :-). I'll let some time (may be a day or 2) for others to possibly come with an answer.
Then I will continue the series with some even harder challenges (formulas + games) on the same theme.
Meanwhile, you can get hints about what's coming by taking a look at MEAN SUMurai (@ Google Play).
Also, don't forget this part of the challenge
"
34 as a mean for 691795
35 as a mean for 59458531
with a least one split to read from right to left.
"
Yeah, I know it's less cool than finding formulas but it requires different skills and qualities.
Best
PS: Latest, more clear version of the challenge below for people who want to try
[CHALLENGE]
Consider that you are given a sequence of n=4 digits
Assign to these digits numbers from 1 to 4.
Below, an enumeration of the all different combinations.
1234
1 2 3 4
1 2 34 -- 1 2 43
1 23 4 -- 1 32 4
12 3 4 -- 21 3 4
12 23 -- 12 32 -- 21 23 -- 21 32
so 11 different splittings for n = 4
Find a formula for the number of combinations, given n.
thx for the welcome :-)
Hm, so, we count 21 1 2 and 12 1 2 as one splitting, but 12 1 2 and 12 2 1 as two, right? And is 3 12 4 a valid splitting if 1234, then?
No, you can't count 21 1 2 and 12 1 2 as one splitting. 2->1 is definitely different from 1->2.
hehe, this is a perfect case-study of why anticipating questions is not always a good thing.
I just confused you guys. Forget totally about same digits stuff and just focus on the following
Consider that you are given a sequence of 4 digits
Assign these digits numbers from 1 to 4.
Below, an enumeration of the all different combinations.
1234
1 2 3 4
1 2 34 -- 1 2 43
1 23 4 -- 1 32 4
12 3 4 -- 21 3 4
12 23 -- 12 32 -- 21 23 -- 21 32
so 11 different splittings for n = 4
*You don't need to care about whether digit 1 is equal or not to digit 2.
*Order is irrelevant. While looking for a formula, you may write 1 2 4 3 instead of
1 2 3 4. It is the same splitting but I highly suggest you stick with the order in
which the digits appear. Otherwise, you may get unnecessarily confused.
[If you want to introduce order, you can do so and it could be interesting
but it is outside the scope of the challenge]
Now, if you really want to see the context of all this, go download the game
Mean Sumurai @ Google Play. This could sound like another shameless plug.
It may well be one :-) but I sincerely think it could help you guys.
Also, don't think this is an easy problem. It is pretty hard and though it has a clean answer
it is not something like Fn+1. It is slightly messier.
Hm, is 21 1 2 a valid splitting of 1221?
hmm 21(12 backwards) 2 1 is a valid splitting and, because the splits do not have to be ordered, so is 21 1 2.
In an attempt of making things more clear, let me propose an alternative formulation of the problem.
The sequence is like a series of dots, and each dot can either be alone or or connected to
its predecessor or successor.
so let's take 1221
(dot1) (dot2) (dot3) (dot4)
21 2 1 corresponds to
(dot1)<-(dot2) (dot3) (dot4)
and because there is no order for the splits
your 21 1 2
(dot1)<-(dot2) (dot4) (dot3)
is also valid, though possibly confusing.
Hi all
I'm from Montreal and I'm a CS/SE(*) researcher with interests in graph theory, NP-hard problems and combinatorial optimisation. During my Ph.D., a minor bug made me accidentally discover "something" that I, probably wrongly, call "number splitting".
Basically, given a number X and a sequence S of digits, can we split S into smaller parts so that the obtained numbers add up (or average) to X?
I don't think this is new territory for maths scholars but I could not find any existing references. May be this is simply not worthy of serious maths investigation. In any case, I made a game out of it: Mean Sumurai on Google Play for people interested [sorry for the plug]. My pseudonym Mean Maths is linked to the game and also my belief that some people out there like their maths to be "mean" and challenging. However, due to family pressure :-D, the game has different difficulty levels. So take a look whatever your actual maths skills ;-P
Best
Mean Maths
(*) Computer Science/Software Engineering
yes. Otherwise, a formula would be much tougher to find.
You can see it the following way. Each digit is assigned a number representing its position in the sequence.
So regardless of the actual digit, a position is always unique. So, if you have 888 as a sequence,
a splitting 8 88[left to right] would correspond to {position1, (position2, position3)}
while a splitting 8 88[right to left] would correspond to {position1, (position3, position2)}
...Are you saying you want each pair counted twice?
You can see it that way, as long as you don't forget that you're looking for the number of combinations.
So, as you probably already know, if you have n pairs in a splitting, it does not mean that it counts for 2*n.
Edit: Just in case: "pair" in what I wrote above refers to 2-digit groups
Hi bobbym
Congrats for the two sums.
I just placed a link to the forum on the game's official website.
meansumz com
All this is a part-time one-man effort, so, as you can see, no fancy stuff.
You look pretty good at number splitting. I have some insane but solvable instances in the game
that you may like. Take a look at the last picture in the download page of
Mean Sumurai in the Google Play Store.
This is one of the toughest instances I solved, in exactly 59 s.
[I integrated in the game the ability to save solved instances.]
Though I'm reasonably good at Maths, I'm far from being a "human calculator".
The secret: a strong mental heuristic that I developed through many many hours of
playing.
Note that the app does contain ads but it mainly allows me to get an idea of how much the game is played.
Right now, the game is played in 22 countries and though there are less than 100 players (I launched the game
4 days ago), I take comfort in the fact that there is always at least one person playing it :-)
Best
Congrats anonimnystefy and bobbym
Your quick answers are impressive, considering that:
- I posted Part II for days in the Mathematics community (~70k) of Google+ and only the first 2 got solved.
- I also posted Part I yesterday before posting it here. still waiting for an answer.
Let's continue this.
__Part III__ (tougher cookie than Part I but I have even tougher ones in case someone manages to solve this.)
You are given a sequence of n digits and allowed to split it by grouping at most 2 consecutive digits together.
Example: 796421 may give the splitting 7 96 4 21
Difference with Part I is that
An orientation is attached to 2-digit splits, meaning they can be read either from left to right or right to left
So 7 96 4 21 (in the above example) may correspond to 7 96 4 21 or 7 96 4 12 or 7 69 4 21 or 7 69 4 12
Whether the two different orientations give the same number or not is not taken into account:
a split 33 (left to right) is different from a split 33 (right to left) because the pairs [2 consecutive number][orientation]
are different
***Find a formula for the number of possible splittings.***
__Part IV__
You are given a number X and a sequence of digits.
Your task is to split the sequence so that the obtained groups
(of at most 2 digits) add up (or average to) X.
You can define your splits from left to right
Examples
24 as a sum for 3381 3 3 81[right to left] --> 3 3 18 --> 24
15 as a mean for 9423 9 4 23[right to left] --> 9 4 32 --> 45/3 = 15
Solve the following:
177 as a sum for 3534477
204 as a sum for 944949165
34 as a mean for 691795
35 as a mean for 59458531
Constraint: Use at least one backward orientation.
In this series, parts with odd numbers will be about formulas
while parts with even numbers will propose some (more and more) challenging
puzzles based on number splitting.
[To the moderator, let me try to convince you not to cut the following :-)
The challenges are taken from the game, which is a free (without any in-app purchases) maths game that may interest people who try to
solve the challenges. I'm a (hobby) app developer but I'm also a computer science Ph.D. with a good day-job. So this is not about money.]
More (diverse and randomly generated) puzzles await you in the Android game
MEAN SUMurai. The game includes an online versus mode, achievements and leaderboards.
If you take a look, you will know that very serious math challenges await you
in the next parts.
Best