The most efficient algorithm to find all possible integer pairs which sum to a given integerWhat algorithm and tools should I use to search a data set for the point nearest to a given point?How do I find the all valid pairings between two sets?All possible pairs of two itemsEfficient way to sum all the primes below $N$ million in MathematicaFind random $n$ combinations of values with a given sumHow to efficiently find all combinations of the letters in an alphabet given a conditionGiven a large binary matrix, find the largest submatrix containing non-zero elementsWhat is the algorithm to find the Up-sets of this set?How can I find the the greatest common divisor with Euclid's algorithm?How to find all prime power factorizations of an integer

What is the maximum amount of diamond in one Minecraft game?

Lie bracket of vector fields in Penrose's abstract index notation

Red and White Squares

How important is it for multiple POVs to run chronologically?

What happens if the limit of 4 billion files was exceeded in an ext4 partition?

What's the big deal about the Nazgûl losing their horses?

PhD: When to quit and move on?

Is this standard Japanese employment negotiations, or am I missing something?

Who is responsible for exterminating cockroaches in house - tenant or landlord?

What causes a fastener to lock?

How can a ban from entering the US be lifted?

What is the difference between an "empty interior" and a "hole" in topology?

Taking my Ph.D. advisor out for dinner after graduation

How would a sea turtle end up on its back?

Do the 26 richest billionaires own as much wealth as the poorest 3.8 billion people?

What do I need to see before Spider-Man: Far From Home?

How to play a D major chord lower than the open E major chord on guitar?

n-level Ouroboros Quine

Why would "dead languages" be the only languages that spells could be written in?

Did Stalin kill all Soviet officers involved in the Winter War?

Is it bad to suddenly introduce another element to your fantasy world a good ways into the story?

Why did moving the mouse cursor cause Windows 95 to run more quickly?

Will Jimmy fall off his platform?

What's the difference between a type and a kind?



The most efficient algorithm to find all possible integer pairs which sum to a given integer


What algorithm and tools should I use to search a data set for the point nearest to a given point?How do I find the all valid pairings between two sets?All possible pairs of two itemsEfficient way to sum all the primes below $N$ million in MathematicaFind random $n$ combinations of values with a given sumHow to efficiently find all combinations of the letters in an alphabet given a conditionGiven a large binary matrix, find the largest submatrix containing non-zero elementsWhat is the algorithm to find the Up-sets of this set?How can I find the the greatest common divisor with Euclid's algorithm?How to find all prime power factorizations of an integer






.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty margin-bottom:0;








3












$begingroup$


I wrote a module in Mathematica which finds all possible pairs of integers from a specified list of integers (which can be negative, zero, or positive) which sum to a specified integer m.



The only limiting assumption this algorithm has is that the user only wishes to get the set of all unique sums which sum to m.



Is there a faster algorithm to do this? I've read that making a Hash table is of complexity O(n). Is my code of time O(n)? If it of time O(n), is it a Hash table, or is it something else? If it is not of time O(n), how efficient is it?



FindTwoIntegersWhoseSumIsM[listOfIntegers_,m_]:=Module[

i,distanceFrom1ToMin,negativeFactor,distance,start,finish,(*Integers*)
sortedList,numberLine,temp,finalList,(*Lists*)
execute(*Boolean*)
,
(*There are possible inputted values of m with a give integer set input which
make the execution of this algorithm unnecessary.*)
execute=True;

sortedList=Sort[DeleteDuplicates[listOfIntegers]];

(*Create a continuous list of integers whose smallest and largest entries is equal
to the smallest and largest entries of the inputted list of integers, respectively.*)
(*Let this list be named numberline.*)

(*:::::Construction of numberline BEGINS::::*)

(*If the listOfIntegers only contains negative integers and possibly zero,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]<=0),

(*If m is positive, there is no reason to proceed.*)

If[m>0,execute=False,
(*If m [Equal] 0 then if two or more zeros are in listOfIntegers, they should be outputted to the user.
Therefore, we write m>0 instead of m[GreaterEqual]0 in the conditional above.*)

(*Otherwise, treat it as if all integers were positive with a few considerations.*)
negativeFactor=-1;
sortedList=Reverse[-sortedList];
If[sortedList[[1]]!=0,
numberLine=Range[sortedList[[Length[sortedList]]]]
,
numberLine=Join[0,Range[sortedList[[Length[sortedList]]]]]
]
]
,
negativeFactor=1;

(*Else If the integer set contains negative and positive integers,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]>0),
numberLine=
Join[
-Range[Abs[sortedList[[1]]],0,-1](*negative integer subset*)
,
Range[sortedList[[Length[sortedList]]]](*positive integer subset*)
]
,(*Else if the integer set contains only whole numbers,*)
If[(sortedList[[1]]==0)&&(sortedList[[Length[sortedList]]]>0),

(*If the list of integers are all positive and m is negative,
there is no reason to proceed.*)
If[m<0,execute=False,(*Otherwise,*)
numberLine=
Join[
0(*zero*)
,
Range[sortedList[[Length[sortedList]]]](*positive integers*)
]
]
,(*Else if the integer set contains only the natural numbers.*)

(*If the list of integers are all positive and m is negative or zero,
there is no reason to proceed.*)
If[m<=0,execute=False,numberLine=Range[Max[sortedList](*positive integers*)]]
]
]
];

(*:::::Construction of numberline ENDS::::*)
(*Print[numberLine];*)


If[execute==False,finalList=$Failed,
(*Mark all numbers which are in numberline but are not in listOfIntegers with a period.

Sort[] will still sort this list of mixed precision of numbers in ascending order.*)
temp=Sort[Join[Complement[numberLine,sortedList]//N,sortedList]];

(*The main idea of the algorithm is to find the point on numberline to begin selecting two number
combinations which sum to m. m is obviously going to be used when that time comes.

Once that point is selected, integers symmetrically equally distant apart from each other
on both sides of this point (number) in numberline are candidates which sum to m.

To avoid going "out of bounds" of numberline (from either attempting to select a value smaller
than the minimum value of numberline or attempting to select a larger value than the maximum
value of numberline, the following is the maximum distance we can use to obtain ALL possible
two integer combinations which sum to m but of which also prevents us from going "out of bounds".)
*)


(*If the numberline we are about to create had a consistent minimum value of 1
then it would not be offset as it is in general.
The following takes this "offset" into account.*)
distanceFrom1ToMin=Abs[1-Min[sortedList]];


distance=
Min[

distanceFrom1ToMin+Floor[negativeFactor*m/2]
,
Length[temp]-(distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1)

];

start=distanceFrom1ToMin+Floor[negativeFactor*m/2]+1;
finish=distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1;

(*With the bound distance established, we are ready to begin selecting numbers from numberline.*)
finalList=;
i=1;
While[i<=distance,
finalList=Append[finalList,temp[[start-i]],temp[[finish+i]]];
i++
];

(*It turns out that for even m the first selected integer combination considered is m/2,m/2.*)
If[(Mod[m,2]==0)&&(MemberQ[finalList,negativeFactor*m/2,negativeFactor*m/2]==True),
(*Should there not be two of m/2 in listOfIntegers, we omit this selected combination.*)
If[Length[Flatten[Position[listOfIntegers,negativeFactor*m/2]]]<2,
finalList=Delete[finalList,Position[finalList,negativeFactor*m/2,negativeFactor*m/2][[1]][[1]]]
]
];

(*We selected all possible number combinations in numberline. However, unless listOfIntegers
is all consecutive integers, we need to omit any selected number combination in which either
of the numbers has a "." to the right of it.*)
finalList=negativeFactor*Sort[Select[finalList,Precision[#]==[Infinity]&]]
];
finalList
]


I did the following tests with the code and got these results. (The first number in the time in second it took to do the computation. But you can of course copy the code and do tests yourself.) I omitted most of the results from the last test because it made my post too large, but you will see that it did the computation in 0.209207 seconds.



As the comments in my algorithm (and the algorithm itself suggests), I broke up the number line into negative integers, zero, and the positive integers. I therefore wrote my tests to address all possible situations.




For the positive (non-zero) integer set.



With positive m such that m is larger than what any two number combination in listOfIntegers could possibly sum to.



m = 100; listOfIntegers = RandomSample[Range[20], 6]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

19, 11, 1, 4, 13, 17

0.0371008,


With positive odd m.



m = 215; listOfIntegers = RandomSample[Range[266], 190]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

119, 175, 7, 123, 42, 173, 15, 56, 233, 41, 9, 156, 130, 196, 183,
65, 102, 109, 177, 161, 230, 105, 91, 103, 146, 47, 234, 133, 88, 68,
169, 197, 46, 198, 108, 263, 205, 129, 4, 157, 245, 210, 203, 78,
172, 128, 138, 61, 262, 159, 148, 45, 225, 239, 72, 74, 151, 34, 36,
5, 106, 77, 223, 116, 8, 2, 11, 54, 124, 87, 221, 213, 171, 93, 53,
19, 40, 30, 95, 215, 39, 140, 49, 158, 94, 38, 28, 247, 84, 75, 257,
33, 163, 132, 69, 211, 193, 222, 114, 240, 32, 149, 167, 135, 107,
115, 101, 100, 166, 144, 251, 253, 224, 154, 48, 44, 26, 181, 259,
81, 6, 70, 122, 255, 189, 235, 112, 110, 174, 85, 147, 117, 18, 209,
66, 121, 155, 206, 207, 212, 98, 113, 254, 214, 178, 111, 227, 165,
204, 231, 194, 20, 176, 150, 162, 241, 243, 199, 90, 55, 127, 191,
12, 185, 242, 125, 265, 25, 1, 250, 201, 168, 76, 134, 266, 82, 10,
92, 143, 217, 126, 218, 182, 220, 153, 164, 216, 238, 67, 14

0.136695, 1, 214, 2, 213, 4, 211, 5, 210, 6, 209, 8,
207, 9, 206, 10, 205, 11, 204, 12, 203, 14, 201, 18,
197, 19, 196, 26, 189, 30, 185, 32, 183, 33, 182, 34,
181, 38, 177, 39, 176, 40, 175, 41, 174, 42, 173, 44,
171, 46, 169, 47, 168, 48, 167, 49, 166, 53, 162, 54,
161, 56, 159, 61, 154, 65, 150, 66, 149, 67, 148, 68,
147, 69, 146, 72, 143, 75, 140, 77, 138, 81, 134, 82,
133, 85, 130, 87, 128, 88, 127, 90, 125, 91, 124, 92,
123, 93, 122, 94, 121, 98, 117, 100, 115, 101,
114, 102, 113, 103, 112, 105, 110, 106, 109, 107, 108


With positive even m.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.00998522, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With positive even m such that listOfIntegers contains two of m/2.



m = 22; listOfIntegers = Append[Range[20], 11]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 11

0.00037181, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12, 11, 11


With positive even m such that listOfIntegers contains one m/2.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.000267311, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With any negative m.



m = -6; listOfIntegers = Range[26]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25, 26

0.000108231, $Failed



For the positive integer set (including 0).



With an even m.



m = 88; listOfIntegers = RandomSample[Join[0, Range[122]], 39]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

121, 69, 120, 56, 36, 55, 17, 114, 7, 59, 32, 4, 20, 79, 92, 62, 50,
89, 13, 70, 113, 75, 76, 80, 108, 53, 83, 95, 0, 85, 86, 77, 10, 54,
48, 66, 104, 100, 35

0.000505232, 13, 75, 32, 56, 35, 53


With an odd m.



m = 57; listOfIntegers = RandomSample[Join[0, Range[82]], 52]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

62, 18, 26, 0, 67, 34, 55, 52, 35, 78, 10, 68, 46, 44, 38, 23, 77,
76, 58, 51, 75, 63, 53, 42, 54, 27, 56, 71, 12, 17, 2, 37, 31, 72,
49, 50, 32, 16, 47, 19, 4, 20, 81, 25, 61, 14, 80, 82, 59, 33, 70, 39

0.000372743, 2, 55, 4, 53, 10, 47, 18, 39, 19, 38, 20,
37, 23, 34, 25, 32, 26, 31



For the negative integer set (including 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-2, -16, -15, -9, -5, -12, -8, -22, -7, -21, -13, -18, -4, -11, -10,
-19, -6, -17, -20

0.000105898, $Failed


With a negative odd m.



m = -17; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-5, -1, -10, -13, -15, -19, -2, 0, -7, -18, -3, -21, -8, -11, -12,
-22, -17, -16, -20

0.000640987, 0, -17, -1, -16, -2, -15, -5, -12, -7, -10


With a negative even m.



m = -26; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -16, -11, -14, -17, -13, -1, -9, -15, -20, -18, -4, -21, 0, -8,
-6, -10, -7, -3

0.000329357, -6, -20, -7, -19, -8, -18, -9, -17, -10,
-16, -11, -15



For the negative integer set (excluding 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-20, -7, -16, -21, -11, -13, -5, -2, -6, -19, -1, -12, -18, -14,
-15, -9, -4, -17, -22

0.000102633, $Failed


With a negative odd m.



m = -27; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-18, -17, -22, -13, -1, -11, -19, -8, -16, -6, -21, -12, -20, -3,
-4, -9, -7, -14, -15

0.000242586, -6, -21, -7, -20, -8, -19, -9, -18, -11,
-16, -12, -15, -13, -14


With a negative even m.



m = -26; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -10, -20, -9, -21, -14, -5, -1, -17, -4, -18, -22, -8, -6, -13,
-3, -2, -12, -15

0.000286438, -4, -22, -5, -21, -6, -20, -8, -18, -9, -17,
-12, -14



For the complete integer set.



With a positive odd m.



m = 15; listOfIntegers = 
RandomSample[Join[-Range[52, 1, -1], 0, Range[52]], 35]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-30, 19, 42, 38, -25, 6, 48, 5, -8, -27, -11, -47, -37, -12, -3,
-34, 50, 11, 10, 18, 7, -15, 51, -22, -26, -2, 33, -35, 34, 39, 44,
-51, -33, -16, -23

0.000468378, -35, 50, -33, 48, -27, 42, -23, 38, -3,
18, 5, 10


With a negative odd m.



m = -7; listOfIntegers = 
RandomSample[Join[-Range[22, 1, -1], 0, Range[22]], 21]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-1, -16, -11, 10, 17, 1, 0, -5, -22, 8, -7, 15, 21, 11, 18, 14, -4,
7, -13, 4, -9

0.000310697, -22, 15, -11, 4, -7, 0


With a positive even m.



m = 36; listOfIntegers = 
RandomSample[Join[-Range[30, 1, -1], 0, Range[30]], 20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

25, -9, -8, 8, 5, -10, -24, 13, 9, -16, -23, -14, -22, -29, 26, 12,
19, 16, -30, 18

0.000289237,


With a negative even m.



m = -34; listOfIntegers = 
RandomSample[Join[-Range[100, 1, -1], 0, Range[100]], 50]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

7, 92, 91, 58, -58, 63, -95, 82, 26, 60, 16, 65, 15, 34, 29, 67, -2,
88, 21, -72, -93, 12, 43, 18, -83, -80, -30, -6, 54, -13, -63, 39,
-55, 9, -78, 5, -16, 52, -24, -82, -18, 2, -90, 37, -60, 80, 57, -22,
-26, 72

0.000726359, -63, 29, -60, 26, -55, 21, -18, -16


With m == 0.



m = 0; listOfIntegers = 
RandomSample[Join[-Range[222, 1, -1], 0, Range[222]], 111]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-215, -8, 186, 153, 17, 83, 149, -45, -18, 14, -161, 6, 84, -41,
-59, -130, 34, -24, -142, -95, -70, -60, -152, 90, -43, 12, -196,
-98, -193, -78, -192, 7, -30, 218, -209, -28, -125, 142, 11, 161,
-143, -135, -212, 134, 1, -177, -100, 2, 63, -180, -50, 79, -129,
-91, 126, 57, -140, -200, 38, -182, -107, -25, -46, -179, -113, 88,
148, 28, 184, -158, 190, -9, -36, -5, 169, 221, -204, -210, 44, 45,
-71, 40, 135, 119, -42, 166, 65, 59, -15, -118, 117, -47, -52, 102,
74, -19, 152, 81, 0, 170, -214, 114, -38, 210, -1, -7, -89, -173,
123, 78, -127

0.00179934, -210, 210, -161, 161, -152, 152, -142,
142, -135, 135, -78, 78, -59, 59, -45, 45, -38,
38, -28, 28, -7, 7, -1, 1


With a large m with a large listOfIntegers.



m = 5311; listOfIntegers = 
RandomSample[Join[-Range[9999, 1, -1], 0, Range[9999]], 8888];
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]

0.209207, -4680, 9991, -4676, 9987, -4664, 9975, -4650,
9961, -4646, 9957, -4645, 9956, -4636, 9947, -4634,
9945, -4633, 9944, -4630, 9941, -4600, 9911, -4599,
9910, -4594, 9905, -4587, 9898, -4574, 9885, -4573,
9884, -4572, 9883, -4566, 9877, -4562, 9873, -4556,
9867, -4549, 9860, -4538, 9849, -4529, 9840, -4517,
9828, -4514, 9825, -4511, 9822, -4504, 9815, -4502,
9813, -4499, 9810, -4497, 9808, -4490, 9801, -4486,
9797, -4485, 9796, -4483, 9794, -4481, 9792, -4478,
9789, -4475, 9786, -4464, 9775, -4463, 9774, -4458,
9769, -4452, 9763, -4443, 9754, -4431, 9742, -4428,
9739, -4427, 9738, -4420, 9731, -4417, 9728, -4407,
9718, -4405, 9716, -4397, 9708, -4394, 9705, -4393,
9704, -4380, 9691, -4377, 9688, -4369, 9680, -4359,
9670, -4356, 9667, -4354, 9665, -4350, 9661, -4349,
9660, -4346, 9657, -4337, 9648, -4332, 9643, -4331,
9642, -4325, 9636, -4323, 9634, -4314, 9625, -4305,
9616, -4293, 9604, -4283, 9594, -4266, 9577, -4246,
9557, -4241, 9552, -4235, 9546, -4231, 9542, -4227,
9538, -4224, 9535, -4222, 9533, -4220, 9531, -4211,
9522, -4203, 9514, -4202, 9513, -4198, 9509, -4196,
9507, -4193, 9504, -4190, 9501, -4181, 9492, -4176,
9487, -4148, 9459, -4138, 9449, -4137, 9448, -4136,
9447, -4127, 9438, -4125, 9436, -4107, 9418, -4086,
9397, -4081, 9392, -4079, 9390, -4078, 9389, -4065,
9376, -4056, 9367, -4041, 9352, -4040, 9351, -4038,
9349, -4035, 9346, -4030, 9341, -4026, 9337, -4020,
9331, -4015, 9326, -4014, 9325, -4010, 9321, -3991,
9302, -3988, 9299, -3984, 9295, -3980, 9291, -3978,
9289, -3977, 9288, -3976, 9287, -3971, 9282, -3970,
9281, -3950, 9261, -3946, 9257, -3938, 9249, -3932,
9243, -3922, 9233, -3920, 9231, -3915, 9226, -3910,
9221, -3909, 9220, -3908, 9219, -3901, 9212, -3900,
9211, -3898, 9209, -3887, 9198, -3885, 9196, -3877,
9188, -3875, 9186, -3869, 9180, -3864, 9175, -3859,
9170, -3854, 9165, -3853, 9164, -3848, 9159, -3839,
9150, -3835, 9146, -3826, 9137, -3821, 9132, -3812,
9123, -3810, 9121, -3807, 9118, -3806, 9117, -3799,
9110, -3797, 9108, -3789, 9100, -3779, 9090, -3777,
9088, -3774, 9085, -3773, 9084, -3769, 9080, -3767,
9078, -3761, 9072, -3751, 9062, -3750, 9061, -3749,
9060, -3748, 9059, -3742, 9053, -3740, 9051, -3731,
9042, -3726, 9037, -3717, 9028, -3715, 9026, -3714,
9025, -3708, 9019, -3704, 9015, -3702, 9013, -3687,
8998, -3677, 8988, -3661, 8972, -3654, 8965, -3653,
8964, -3649, 8960, -3641, 8952, -3635, 8946, -3622,
8933, -3615, 8926, -3610, 8921, -3607, 8918, -3601,
8912, -3597, 8908, -3592, 8903, -3586, 8897, ... , 2594, 2717, 2598, 2713, 2599, 2712, 2603,
2708, 2607, 2704, 2617, 2694, 2619, 2692, 2633,
2678, 2634, 2677, 2643, 2668, 2644, 2667, 2648,
2663, 2650, 2661









share|improve this question











$endgroup$











  • $begingroup$
    The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
    $endgroup$
    – Henrik Schumacher
    Mar 25 at 18:04










  • $begingroup$
    You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
    $endgroup$
    – MikeY
    Mar 25 at 18:37










  • $begingroup$
    According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:58










  • $begingroup$
    Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
    $endgroup$
    – Daniel Lichtblau
    Mar 27 at 23:15

















3












$begingroup$


I wrote a module in Mathematica which finds all possible pairs of integers from a specified list of integers (which can be negative, zero, or positive) which sum to a specified integer m.



The only limiting assumption this algorithm has is that the user only wishes to get the set of all unique sums which sum to m.



Is there a faster algorithm to do this? I've read that making a Hash table is of complexity O(n). Is my code of time O(n)? If it of time O(n), is it a Hash table, or is it something else? If it is not of time O(n), how efficient is it?



FindTwoIntegersWhoseSumIsM[listOfIntegers_,m_]:=Module[

i,distanceFrom1ToMin,negativeFactor,distance,start,finish,(*Integers*)
sortedList,numberLine,temp,finalList,(*Lists*)
execute(*Boolean*)
,
(*There are possible inputted values of m with a give integer set input which
make the execution of this algorithm unnecessary.*)
execute=True;

sortedList=Sort[DeleteDuplicates[listOfIntegers]];

(*Create a continuous list of integers whose smallest and largest entries is equal
to the smallest and largest entries of the inputted list of integers, respectively.*)
(*Let this list be named numberline.*)

(*:::::Construction of numberline BEGINS::::*)

(*If the listOfIntegers only contains negative integers and possibly zero,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]<=0),

(*If m is positive, there is no reason to proceed.*)

If[m>0,execute=False,
(*If m [Equal] 0 then if two or more zeros are in listOfIntegers, they should be outputted to the user.
Therefore, we write m>0 instead of m[GreaterEqual]0 in the conditional above.*)

(*Otherwise, treat it as if all integers were positive with a few considerations.*)
negativeFactor=-1;
sortedList=Reverse[-sortedList];
If[sortedList[[1]]!=0,
numberLine=Range[sortedList[[Length[sortedList]]]]
,
numberLine=Join[0,Range[sortedList[[Length[sortedList]]]]]
]
]
,
negativeFactor=1;

(*Else If the integer set contains negative and positive integers,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]>0),
numberLine=
Join[
-Range[Abs[sortedList[[1]]],0,-1](*negative integer subset*)
,
Range[sortedList[[Length[sortedList]]]](*positive integer subset*)
]
,(*Else if the integer set contains only whole numbers,*)
If[(sortedList[[1]]==0)&&(sortedList[[Length[sortedList]]]>0),

(*If the list of integers are all positive and m is negative,
there is no reason to proceed.*)
If[m<0,execute=False,(*Otherwise,*)
numberLine=
Join[
0(*zero*)
,
Range[sortedList[[Length[sortedList]]]](*positive integers*)
]
]
,(*Else if the integer set contains only the natural numbers.*)

(*If the list of integers are all positive and m is negative or zero,
there is no reason to proceed.*)
If[m<=0,execute=False,numberLine=Range[Max[sortedList](*positive integers*)]]
]
]
];

(*:::::Construction of numberline ENDS::::*)
(*Print[numberLine];*)


If[execute==False,finalList=$Failed,
(*Mark all numbers which are in numberline but are not in listOfIntegers with a period.

Sort[] will still sort this list of mixed precision of numbers in ascending order.*)
temp=Sort[Join[Complement[numberLine,sortedList]//N,sortedList]];

(*The main idea of the algorithm is to find the point on numberline to begin selecting two number
combinations which sum to m. m is obviously going to be used when that time comes.

Once that point is selected, integers symmetrically equally distant apart from each other
on both sides of this point (number) in numberline are candidates which sum to m.

To avoid going "out of bounds" of numberline (from either attempting to select a value smaller
than the minimum value of numberline or attempting to select a larger value than the maximum
value of numberline, the following is the maximum distance we can use to obtain ALL possible
two integer combinations which sum to m but of which also prevents us from going "out of bounds".)
*)


(*If the numberline we are about to create had a consistent minimum value of 1
then it would not be offset as it is in general.
The following takes this "offset" into account.*)
distanceFrom1ToMin=Abs[1-Min[sortedList]];


distance=
Min[

distanceFrom1ToMin+Floor[negativeFactor*m/2]
,
Length[temp]-(distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1)

];

start=distanceFrom1ToMin+Floor[negativeFactor*m/2]+1;
finish=distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1;

(*With the bound distance established, we are ready to begin selecting numbers from numberline.*)
finalList=;
i=1;
While[i<=distance,
finalList=Append[finalList,temp[[start-i]],temp[[finish+i]]];
i++
];

(*It turns out that for even m the first selected integer combination considered is m/2,m/2.*)
If[(Mod[m,2]==0)&&(MemberQ[finalList,negativeFactor*m/2,negativeFactor*m/2]==True),
(*Should there not be two of m/2 in listOfIntegers, we omit this selected combination.*)
If[Length[Flatten[Position[listOfIntegers,negativeFactor*m/2]]]<2,
finalList=Delete[finalList,Position[finalList,negativeFactor*m/2,negativeFactor*m/2][[1]][[1]]]
]
];

(*We selected all possible number combinations in numberline. However, unless listOfIntegers
is all consecutive integers, we need to omit any selected number combination in which either
of the numbers has a "." to the right of it.*)
finalList=negativeFactor*Sort[Select[finalList,Precision[#]==[Infinity]&]]
];
finalList
]


I did the following tests with the code and got these results. (The first number in the time in second it took to do the computation. But you can of course copy the code and do tests yourself.) I omitted most of the results from the last test because it made my post too large, but you will see that it did the computation in 0.209207 seconds.



As the comments in my algorithm (and the algorithm itself suggests), I broke up the number line into negative integers, zero, and the positive integers. I therefore wrote my tests to address all possible situations.




For the positive (non-zero) integer set.



With positive m such that m is larger than what any two number combination in listOfIntegers could possibly sum to.



m = 100; listOfIntegers = RandomSample[Range[20], 6]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

19, 11, 1, 4, 13, 17

0.0371008,


With positive odd m.



m = 215; listOfIntegers = RandomSample[Range[266], 190]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

119, 175, 7, 123, 42, 173, 15, 56, 233, 41, 9, 156, 130, 196, 183,
65, 102, 109, 177, 161, 230, 105, 91, 103, 146, 47, 234, 133, 88, 68,
169, 197, 46, 198, 108, 263, 205, 129, 4, 157, 245, 210, 203, 78,
172, 128, 138, 61, 262, 159, 148, 45, 225, 239, 72, 74, 151, 34, 36,
5, 106, 77, 223, 116, 8, 2, 11, 54, 124, 87, 221, 213, 171, 93, 53,
19, 40, 30, 95, 215, 39, 140, 49, 158, 94, 38, 28, 247, 84, 75, 257,
33, 163, 132, 69, 211, 193, 222, 114, 240, 32, 149, 167, 135, 107,
115, 101, 100, 166, 144, 251, 253, 224, 154, 48, 44, 26, 181, 259,
81, 6, 70, 122, 255, 189, 235, 112, 110, 174, 85, 147, 117, 18, 209,
66, 121, 155, 206, 207, 212, 98, 113, 254, 214, 178, 111, 227, 165,
204, 231, 194, 20, 176, 150, 162, 241, 243, 199, 90, 55, 127, 191,
12, 185, 242, 125, 265, 25, 1, 250, 201, 168, 76, 134, 266, 82, 10,
92, 143, 217, 126, 218, 182, 220, 153, 164, 216, 238, 67, 14

0.136695, 1, 214, 2, 213, 4, 211, 5, 210, 6, 209, 8,
207, 9, 206, 10, 205, 11, 204, 12, 203, 14, 201, 18,
197, 19, 196, 26, 189, 30, 185, 32, 183, 33, 182, 34,
181, 38, 177, 39, 176, 40, 175, 41, 174, 42, 173, 44,
171, 46, 169, 47, 168, 48, 167, 49, 166, 53, 162, 54,
161, 56, 159, 61, 154, 65, 150, 66, 149, 67, 148, 68,
147, 69, 146, 72, 143, 75, 140, 77, 138, 81, 134, 82,
133, 85, 130, 87, 128, 88, 127, 90, 125, 91, 124, 92,
123, 93, 122, 94, 121, 98, 117, 100, 115, 101,
114, 102, 113, 103, 112, 105, 110, 106, 109, 107, 108


With positive even m.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.00998522, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With positive even m such that listOfIntegers contains two of m/2.



m = 22; listOfIntegers = Append[Range[20], 11]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 11

0.00037181, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12, 11, 11


With positive even m such that listOfIntegers contains one m/2.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.000267311, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With any negative m.



m = -6; listOfIntegers = Range[26]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25, 26

0.000108231, $Failed



For the positive integer set (including 0).



With an even m.



m = 88; listOfIntegers = RandomSample[Join[0, Range[122]], 39]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

121, 69, 120, 56, 36, 55, 17, 114, 7, 59, 32, 4, 20, 79, 92, 62, 50,
89, 13, 70, 113, 75, 76, 80, 108, 53, 83, 95, 0, 85, 86, 77, 10, 54,
48, 66, 104, 100, 35

0.000505232, 13, 75, 32, 56, 35, 53


With an odd m.



m = 57; listOfIntegers = RandomSample[Join[0, Range[82]], 52]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

62, 18, 26, 0, 67, 34, 55, 52, 35, 78, 10, 68, 46, 44, 38, 23, 77,
76, 58, 51, 75, 63, 53, 42, 54, 27, 56, 71, 12, 17, 2, 37, 31, 72,
49, 50, 32, 16, 47, 19, 4, 20, 81, 25, 61, 14, 80, 82, 59, 33, 70, 39

0.000372743, 2, 55, 4, 53, 10, 47, 18, 39, 19, 38, 20,
37, 23, 34, 25, 32, 26, 31



For the negative integer set (including 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-2, -16, -15, -9, -5, -12, -8, -22, -7, -21, -13, -18, -4, -11, -10,
-19, -6, -17, -20

0.000105898, $Failed


With a negative odd m.



m = -17; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-5, -1, -10, -13, -15, -19, -2, 0, -7, -18, -3, -21, -8, -11, -12,
-22, -17, -16, -20

0.000640987, 0, -17, -1, -16, -2, -15, -5, -12, -7, -10


With a negative even m.



m = -26; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -16, -11, -14, -17, -13, -1, -9, -15, -20, -18, -4, -21, 0, -8,
-6, -10, -7, -3

0.000329357, -6, -20, -7, -19, -8, -18, -9, -17, -10,
-16, -11, -15



For the negative integer set (excluding 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-20, -7, -16, -21, -11, -13, -5, -2, -6, -19, -1, -12, -18, -14,
-15, -9, -4, -17, -22

0.000102633, $Failed


With a negative odd m.



m = -27; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-18, -17, -22, -13, -1, -11, -19, -8, -16, -6, -21, -12, -20, -3,
-4, -9, -7, -14, -15

0.000242586, -6, -21, -7, -20, -8, -19, -9, -18, -11,
-16, -12, -15, -13, -14


With a negative even m.



m = -26; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -10, -20, -9, -21, -14, -5, -1, -17, -4, -18, -22, -8, -6, -13,
-3, -2, -12, -15

0.000286438, -4, -22, -5, -21, -6, -20, -8, -18, -9, -17,
-12, -14



For the complete integer set.



With a positive odd m.



m = 15; listOfIntegers = 
RandomSample[Join[-Range[52, 1, -1], 0, Range[52]], 35]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-30, 19, 42, 38, -25, 6, 48, 5, -8, -27, -11, -47, -37, -12, -3,
-34, 50, 11, 10, 18, 7, -15, 51, -22, -26, -2, 33, -35, 34, 39, 44,
-51, -33, -16, -23

0.000468378, -35, 50, -33, 48, -27, 42, -23, 38, -3,
18, 5, 10


With a negative odd m.



m = -7; listOfIntegers = 
RandomSample[Join[-Range[22, 1, -1], 0, Range[22]], 21]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-1, -16, -11, 10, 17, 1, 0, -5, -22, 8, -7, 15, 21, 11, 18, 14, -4,
7, -13, 4, -9

0.000310697, -22, 15, -11, 4, -7, 0


With a positive even m.



m = 36; listOfIntegers = 
RandomSample[Join[-Range[30, 1, -1], 0, Range[30]], 20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

25, -9, -8, 8, 5, -10, -24, 13, 9, -16, -23, -14, -22, -29, 26, 12,
19, 16, -30, 18

0.000289237,


With a negative even m.



m = -34; listOfIntegers = 
RandomSample[Join[-Range[100, 1, -1], 0, Range[100]], 50]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

7, 92, 91, 58, -58, 63, -95, 82, 26, 60, 16, 65, 15, 34, 29, 67, -2,
88, 21, -72, -93, 12, 43, 18, -83, -80, -30, -6, 54, -13, -63, 39,
-55, 9, -78, 5, -16, 52, -24, -82, -18, 2, -90, 37, -60, 80, 57, -22,
-26, 72

0.000726359, -63, 29, -60, 26, -55, 21, -18, -16


With m == 0.



m = 0; listOfIntegers = 
RandomSample[Join[-Range[222, 1, -1], 0, Range[222]], 111]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-215, -8, 186, 153, 17, 83, 149, -45, -18, 14, -161, 6, 84, -41,
-59, -130, 34, -24, -142, -95, -70, -60, -152, 90, -43, 12, -196,
-98, -193, -78, -192, 7, -30, 218, -209, -28, -125, 142, 11, 161,
-143, -135, -212, 134, 1, -177, -100, 2, 63, -180, -50, 79, -129,
-91, 126, 57, -140, -200, 38, -182, -107, -25, -46, -179, -113, 88,
148, 28, 184, -158, 190, -9, -36, -5, 169, 221, -204, -210, 44, 45,
-71, 40, 135, 119, -42, 166, 65, 59, -15, -118, 117, -47, -52, 102,
74, -19, 152, 81, 0, 170, -214, 114, -38, 210, -1, -7, -89, -173,
123, 78, -127

0.00179934, -210, 210, -161, 161, -152, 152, -142,
142, -135, 135, -78, 78, -59, 59, -45, 45, -38,
38, -28, 28, -7, 7, -1, 1


With a large m with a large listOfIntegers.



m = 5311; listOfIntegers = 
RandomSample[Join[-Range[9999, 1, -1], 0, Range[9999]], 8888];
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]

0.209207, -4680, 9991, -4676, 9987, -4664, 9975, -4650,
9961, -4646, 9957, -4645, 9956, -4636, 9947, -4634,
9945, -4633, 9944, -4630, 9941, -4600, 9911, -4599,
9910, -4594, 9905, -4587, 9898, -4574, 9885, -4573,
9884, -4572, 9883, -4566, 9877, -4562, 9873, -4556,
9867, -4549, 9860, -4538, 9849, -4529, 9840, -4517,
9828, -4514, 9825, -4511, 9822, -4504, 9815, -4502,
9813, -4499, 9810, -4497, 9808, -4490, 9801, -4486,
9797, -4485, 9796, -4483, 9794, -4481, 9792, -4478,
9789, -4475, 9786, -4464, 9775, -4463, 9774, -4458,
9769, -4452, 9763, -4443, 9754, -4431, 9742, -4428,
9739, -4427, 9738, -4420, 9731, -4417, 9728, -4407,
9718, -4405, 9716, -4397, 9708, -4394, 9705, -4393,
9704, -4380, 9691, -4377, 9688, -4369, 9680, -4359,
9670, -4356, 9667, -4354, 9665, -4350, 9661, -4349,
9660, -4346, 9657, -4337, 9648, -4332, 9643, -4331,
9642, -4325, 9636, -4323, 9634, -4314, 9625, -4305,
9616, -4293, 9604, -4283, 9594, -4266, 9577, -4246,
9557, -4241, 9552, -4235, 9546, -4231, 9542, -4227,
9538, -4224, 9535, -4222, 9533, -4220, 9531, -4211,
9522, -4203, 9514, -4202, 9513, -4198, 9509, -4196,
9507, -4193, 9504, -4190, 9501, -4181, 9492, -4176,
9487, -4148, 9459, -4138, 9449, -4137, 9448, -4136,
9447, -4127, 9438, -4125, 9436, -4107, 9418, -4086,
9397, -4081, 9392, -4079, 9390, -4078, 9389, -4065,
9376, -4056, 9367, -4041, 9352, -4040, 9351, -4038,
9349, -4035, 9346, -4030, 9341, -4026, 9337, -4020,
9331, -4015, 9326, -4014, 9325, -4010, 9321, -3991,
9302, -3988, 9299, -3984, 9295, -3980, 9291, -3978,
9289, -3977, 9288, -3976, 9287, -3971, 9282, -3970,
9281, -3950, 9261, -3946, 9257, -3938, 9249, -3932,
9243, -3922, 9233, -3920, 9231, -3915, 9226, -3910,
9221, -3909, 9220, -3908, 9219, -3901, 9212, -3900,
9211, -3898, 9209, -3887, 9198, -3885, 9196, -3877,
9188, -3875, 9186, -3869, 9180, -3864, 9175, -3859,
9170, -3854, 9165, -3853, 9164, -3848, 9159, -3839,
9150, -3835, 9146, -3826, 9137, -3821, 9132, -3812,
9123, -3810, 9121, -3807, 9118, -3806, 9117, -3799,
9110, -3797, 9108, -3789, 9100, -3779, 9090, -3777,
9088, -3774, 9085, -3773, 9084, -3769, 9080, -3767,
9078, -3761, 9072, -3751, 9062, -3750, 9061, -3749,
9060, -3748, 9059, -3742, 9053, -3740, 9051, -3731,
9042, -3726, 9037, -3717, 9028, -3715, 9026, -3714,
9025, -3708, 9019, -3704, 9015, -3702, 9013, -3687,
8998, -3677, 8988, -3661, 8972, -3654, 8965, -3653,
8964, -3649, 8960, -3641, 8952, -3635, 8946, -3622,
8933, -3615, 8926, -3610, 8921, -3607, 8918, -3601,
8912, -3597, 8908, -3592, 8903, -3586, 8897, ... , 2594, 2717, 2598, 2713, 2599, 2712, 2603,
2708, 2607, 2704, 2617, 2694, 2619, 2692, 2633,
2678, 2634, 2677, 2643, 2668, 2644, 2667, 2648,
2663, 2650, 2661









share|improve this question











$endgroup$











  • $begingroup$
    The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
    $endgroup$
    – Henrik Schumacher
    Mar 25 at 18:04










  • $begingroup$
    You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
    $endgroup$
    – MikeY
    Mar 25 at 18:37










  • $begingroup$
    According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:58










  • $begingroup$
    Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
    $endgroup$
    – Daniel Lichtblau
    Mar 27 at 23:15













3












3








3





$begingroup$


I wrote a module in Mathematica which finds all possible pairs of integers from a specified list of integers (which can be negative, zero, or positive) which sum to a specified integer m.



The only limiting assumption this algorithm has is that the user only wishes to get the set of all unique sums which sum to m.



Is there a faster algorithm to do this? I've read that making a Hash table is of complexity O(n). Is my code of time O(n)? If it of time O(n), is it a Hash table, or is it something else? If it is not of time O(n), how efficient is it?



FindTwoIntegersWhoseSumIsM[listOfIntegers_,m_]:=Module[

i,distanceFrom1ToMin,negativeFactor,distance,start,finish,(*Integers*)
sortedList,numberLine,temp,finalList,(*Lists*)
execute(*Boolean*)
,
(*There are possible inputted values of m with a give integer set input which
make the execution of this algorithm unnecessary.*)
execute=True;

sortedList=Sort[DeleteDuplicates[listOfIntegers]];

(*Create a continuous list of integers whose smallest and largest entries is equal
to the smallest and largest entries of the inputted list of integers, respectively.*)
(*Let this list be named numberline.*)

(*:::::Construction of numberline BEGINS::::*)

(*If the listOfIntegers only contains negative integers and possibly zero,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]<=0),

(*If m is positive, there is no reason to proceed.*)

If[m>0,execute=False,
(*If m [Equal] 0 then if two or more zeros are in listOfIntegers, they should be outputted to the user.
Therefore, we write m>0 instead of m[GreaterEqual]0 in the conditional above.*)

(*Otherwise, treat it as if all integers were positive with a few considerations.*)
negativeFactor=-1;
sortedList=Reverse[-sortedList];
If[sortedList[[1]]!=0,
numberLine=Range[sortedList[[Length[sortedList]]]]
,
numberLine=Join[0,Range[sortedList[[Length[sortedList]]]]]
]
]
,
negativeFactor=1;

(*Else If the integer set contains negative and positive integers,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]>0),
numberLine=
Join[
-Range[Abs[sortedList[[1]]],0,-1](*negative integer subset*)
,
Range[sortedList[[Length[sortedList]]]](*positive integer subset*)
]
,(*Else if the integer set contains only whole numbers,*)
If[(sortedList[[1]]==0)&&(sortedList[[Length[sortedList]]]>0),

(*If the list of integers are all positive and m is negative,
there is no reason to proceed.*)
If[m<0,execute=False,(*Otherwise,*)
numberLine=
Join[
0(*zero*)
,
Range[sortedList[[Length[sortedList]]]](*positive integers*)
]
]
,(*Else if the integer set contains only the natural numbers.*)

(*If the list of integers are all positive and m is negative or zero,
there is no reason to proceed.*)
If[m<=0,execute=False,numberLine=Range[Max[sortedList](*positive integers*)]]
]
]
];

(*:::::Construction of numberline ENDS::::*)
(*Print[numberLine];*)


If[execute==False,finalList=$Failed,
(*Mark all numbers which are in numberline but are not in listOfIntegers with a period.

Sort[] will still sort this list of mixed precision of numbers in ascending order.*)
temp=Sort[Join[Complement[numberLine,sortedList]//N,sortedList]];

(*The main idea of the algorithm is to find the point on numberline to begin selecting two number
combinations which sum to m. m is obviously going to be used when that time comes.

Once that point is selected, integers symmetrically equally distant apart from each other
on both sides of this point (number) in numberline are candidates which sum to m.

To avoid going "out of bounds" of numberline (from either attempting to select a value smaller
than the minimum value of numberline or attempting to select a larger value than the maximum
value of numberline, the following is the maximum distance we can use to obtain ALL possible
two integer combinations which sum to m but of which also prevents us from going "out of bounds".)
*)


(*If the numberline we are about to create had a consistent minimum value of 1
then it would not be offset as it is in general.
The following takes this "offset" into account.*)
distanceFrom1ToMin=Abs[1-Min[sortedList]];


distance=
Min[

distanceFrom1ToMin+Floor[negativeFactor*m/2]
,
Length[temp]-(distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1)

];

start=distanceFrom1ToMin+Floor[negativeFactor*m/2]+1;
finish=distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1;

(*With the bound distance established, we are ready to begin selecting numbers from numberline.*)
finalList=;
i=1;
While[i<=distance,
finalList=Append[finalList,temp[[start-i]],temp[[finish+i]]];
i++
];

(*It turns out that for even m the first selected integer combination considered is m/2,m/2.*)
If[(Mod[m,2]==0)&&(MemberQ[finalList,negativeFactor*m/2,negativeFactor*m/2]==True),
(*Should there not be two of m/2 in listOfIntegers, we omit this selected combination.*)
If[Length[Flatten[Position[listOfIntegers,negativeFactor*m/2]]]<2,
finalList=Delete[finalList,Position[finalList,negativeFactor*m/2,negativeFactor*m/2][[1]][[1]]]
]
];

(*We selected all possible number combinations in numberline. However, unless listOfIntegers
is all consecutive integers, we need to omit any selected number combination in which either
of the numbers has a "." to the right of it.*)
finalList=negativeFactor*Sort[Select[finalList,Precision[#]==[Infinity]&]]
];
finalList
]


I did the following tests with the code and got these results. (The first number in the time in second it took to do the computation. But you can of course copy the code and do tests yourself.) I omitted most of the results from the last test because it made my post too large, but you will see that it did the computation in 0.209207 seconds.



As the comments in my algorithm (and the algorithm itself suggests), I broke up the number line into negative integers, zero, and the positive integers. I therefore wrote my tests to address all possible situations.




For the positive (non-zero) integer set.



With positive m such that m is larger than what any two number combination in listOfIntegers could possibly sum to.



m = 100; listOfIntegers = RandomSample[Range[20], 6]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

19, 11, 1, 4, 13, 17

0.0371008,


With positive odd m.



m = 215; listOfIntegers = RandomSample[Range[266], 190]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

119, 175, 7, 123, 42, 173, 15, 56, 233, 41, 9, 156, 130, 196, 183,
65, 102, 109, 177, 161, 230, 105, 91, 103, 146, 47, 234, 133, 88, 68,
169, 197, 46, 198, 108, 263, 205, 129, 4, 157, 245, 210, 203, 78,
172, 128, 138, 61, 262, 159, 148, 45, 225, 239, 72, 74, 151, 34, 36,
5, 106, 77, 223, 116, 8, 2, 11, 54, 124, 87, 221, 213, 171, 93, 53,
19, 40, 30, 95, 215, 39, 140, 49, 158, 94, 38, 28, 247, 84, 75, 257,
33, 163, 132, 69, 211, 193, 222, 114, 240, 32, 149, 167, 135, 107,
115, 101, 100, 166, 144, 251, 253, 224, 154, 48, 44, 26, 181, 259,
81, 6, 70, 122, 255, 189, 235, 112, 110, 174, 85, 147, 117, 18, 209,
66, 121, 155, 206, 207, 212, 98, 113, 254, 214, 178, 111, 227, 165,
204, 231, 194, 20, 176, 150, 162, 241, 243, 199, 90, 55, 127, 191,
12, 185, 242, 125, 265, 25, 1, 250, 201, 168, 76, 134, 266, 82, 10,
92, 143, 217, 126, 218, 182, 220, 153, 164, 216, 238, 67, 14

0.136695, 1, 214, 2, 213, 4, 211, 5, 210, 6, 209, 8,
207, 9, 206, 10, 205, 11, 204, 12, 203, 14, 201, 18,
197, 19, 196, 26, 189, 30, 185, 32, 183, 33, 182, 34,
181, 38, 177, 39, 176, 40, 175, 41, 174, 42, 173, 44,
171, 46, 169, 47, 168, 48, 167, 49, 166, 53, 162, 54,
161, 56, 159, 61, 154, 65, 150, 66, 149, 67, 148, 68,
147, 69, 146, 72, 143, 75, 140, 77, 138, 81, 134, 82,
133, 85, 130, 87, 128, 88, 127, 90, 125, 91, 124, 92,
123, 93, 122, 94, 121, 98, 117, 100, 115, 101,
114, 102, 113, 103, 112, 105, 110, 106, 109, 107, 108


With positive even m.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.00998522, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With positive even m such that listOfIntegers contains two of m/2.



m = 22; listOfIntegers = Append[Range[20], 11]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 11

0.00037181, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12, 11, 11


With positive even m such that listOfIntegers contains one m/2.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.000267311, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With any negative m.



m = -6; listOfIntegers = Range[26]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25, 26

0.000108231, $Failed



For the positive integer set (including 0).



With an even m.



m = 88; listOfIntegers = RandomSample[Join[0, Range[122]], 39]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

121, 69, 120, 56, 36, 55, 17, 114, 7, 59, 32, 4, 20, 79, 92, 62, 50,
89, 13, 70, 113, 75, 76, 80, 108, 53, 83, 95, 0, 85, 86, 77, 10, 54,
48, 66, 104, 100, 35

0.000505232, 13, 75, 32, 56, 35, 53


With an odd m.



m = 57; listOfIntegers = RandomSample[Join[0, Range[82]], 52]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

62, 18, 26, 0, 67, 34, 55, 52, 35, 78, 10, 68, 46, 44, 38, 23, 77,
76, 58, 51, 75, 63, 53, 42, 54, 27, 56, 71, 12, 17, 2, 37, 31, 72,
49, 50, 32, 16, 47, 19, 4, 20, 81, 25, 61, 14, 80, 82, 59, 33, 70, 39

0.000372743, 2, 55, 4, 53, 10, 47, 18, 39, 19, 38, 20,
37, 23, 34, 25, 32, 26, 31



For the negative integer set (including 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-2, -16, -15, -9, -5, -12, -8, -22, -7, -21, -13, -18, -4, -11, -10,
-19, -6, -17, -20

0.000105898, $Failed


With a negative odd m.



m = -17; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-5, -1, -10, -13, -15, -19, -2, 0, -7, -18, -3, -21, -8, -11, -12,
-22, -17, -16, -20

0.000640987, 0, -17, -1, -16, -2, -15, -5, -12, -7, -10


With a negative even m.



m = -26; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -16, -11, -14, -17, -13, -1, -9, -15, -20, -18, -4, -21, 0, -8,
-6, -10, -7, -3

0.000329357, -6, -20, -7, -19, -8, -18, -9, -17, -10,
-16, -11, -15



For the negative integer set (excluding 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-20, -7, -16, -21, -11, -13, -5, -2, -6, -19, -1, -12, -18, -14,
-15, -9, -4, -17, -22

0.000102633, $Failed


With a negative odd m.



m = -27; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-18, -17, -22, -13, -1, -11, -19, -8, -16, -6, -21, -12, -20, -3,
-4, -9, -7, -14, -15

0.000242586, -6, -21, -7, -20, -8, -19, -9, -18, -11,
-16, -12, -15, -13, -14


With a negative even m.



m = -26; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -10, -20, -9, -21, -14, -5, -1, -17, -4, -18, -22, -8, -6, -13,
-3, -2, -12, -15

0.000286438, -4, -22, -5, -21, -6, -20, -8, -18, -9, -17,
-12, -14



For the complete integer set.



With a positive odd m.



m = 15; listOfIntegers = 
RandomSample[Join[-Range[52, 1, -1], 0, Range[52]], 35]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-30, 19, 42, 38, -25, 6, 48, 5, -8, -27, -11, -47, -37, -12, -3,
-34, 50, 11, 10, 18, 7, -15, 51, -22, -26, -2, 33, -35, 34, 39, 44,
-51, -33, -16, -23

0.000468378, -35, 50, -33, 48, -27, 42, -23, 38, -3,
18, 5, 10


With a negative odd m.



m = -7; listOfIntegers = 
RandomSample[Join[-Range[22, 1, -1], 0, Range[22]], 21]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-1, -16, -11, 10, 17, 1, 0, -5, -22, 8, -7, 15, 21, 11, 18, 14, -4,
7, -13, 4, -9

0.000310697, -22, 15, -11, 4, -7, 0


With a positive even m.



m = 36; listOfIntegers = 
RandomSample[Join[-Range[30, 1, -1], 0, Range[30]], 20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

25, -9, -8, 8, 5, -10, -24, 13, 9, -16, -23, -14, -22, -29, 26, 12,
19, 16, -30, 18

0.000289237,


With a negative even m.



m = -34; listOfIntegers = 
RandomSample[Join[-Range[100, 1, -1], 0, Range[100]], 50]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

7, 92, 91, 58, -58, 63, -95, 82, 26, 60, 16, 65, 15, 34, 29, 67, -2,
88, 21, -72, -93, 12, 43, 18, -83, -80, -30, -6, 54, -13, -63, 39,
-55, 9, -78, 5, -16, 52, -24, -82, -18, 2, -90, 37, -60, 80, 57, -22,
-26, 72

0.000726359, -63, 29, -60, 26, -55, 21, -18, -16


With m == 0.



m = 0; listOfIntegers = 
RandomSample[Join[-Range[222, 1, -1], 0, Range[222]], 111]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-215, -8, 186, 153, 17, 83, 149, -45, -18, 14, -161, 6, 84, -41,
-59, -130, 34, -24, -142, -95, -70, -60, -152, 90, -43, 12, -196,
-98, -193, -78, -192, 7, -30, 218, -209, -28, -125, 142, 11, 161,
-143, -135, -212, 134, 1, -177, -100, 2, 63, -180, -50, 79, -129,
-91, 126, 57, -140, -200, 38, -182, -107, -25, -46, -179, -113, 88,
148, 28, 184, -158, 190, -9, -36, -5, 169, 221, -204, -210, 44, 45,
-71, 40, 135, 119, -42, 166, 65, 59, -15, -118, 117, -47, -52, 102,
74, -19, 152, 81, 0, 170, -214, 114, -38, 210, -1, -7, -89, -173,
123, 78, -127

0.00179934, -210, 210, -161, 161, -152, 152, -142,
142, -135, 135, -78, 78, -59, 59, -45, 45, -38,
38, -28, 28, -7, 7, -1, 1


With a large m with a large listOfIntegers.



m = 5311; listOfIntegers = 
RandomSample[Join[-Range[9999, 1, -1], 0, Range[9999]], 8888];
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]

0.209207, -4680, 9991, -4676, 9987, -4664, 9975, -4650,
9961, -4646, 9957, -4645, 9956, -4636, 9947, -4634,
9945, -4633, 9944, -4630, 9941, -4600, 9911, -4599,
9910, -4594, 9905, -4587, 9898, -4574, 9885, -4573,
9884, -4572, 9883, -4566, 9877, -4562, 9873, -4556,
9867, -4549, 9860, -4538, 9849, -4529, 9840, -4517,
9828, -4514, 9825, -4511, 9822, -4504, 9815, -4502,
9813, -4499, 9810, -4497, 9808, -4490, 9801, -4486,
9797, -4485, 9796, -4483, 9794, -4481, 9792, -4478,
9789, -4475, 9786, -4464, 9775, -4463, 9774, -4458,
9769, -4452, 9763, -4443, 9754, -4431, 9742, -4428,
9739, -4427, 9738, -4420, 9731, -4417, 9728, -4407,
9718, -4405, 9716, -4397, 9708, -4394, 9705, -4393,
9704, -4380, 9691, -4377, 9688, -4369, 9680, -4359,
9670, -4356, 9667, -4354, 9665, -4350, 9661, -4349,
9660, -4346, 9657, -4337, 9648, -4332, 9643, -4331,
9642, -4325, 9636, -4323, 9634, -4314, 9625, -4305,
9616, -4293, 9604, -4283, 9594, -4266, 9577, -4246,
9557, -4241, 9552, -4235, 9546, -4231, 9542, -4227,
9538, -4224, 9535, -4222, 9533, -4220, 9531, -4211,
9522, -4203, 9514, -4202, 9513, -4198, 9509, -4196,
9507, -4193, 9504, -4190, 9501, -4181, 9492, -4176,
9487, -4148, 9459, -4138, 9449, -4137, 9448, -4136,
9447, -4127, 9438, -4125, 9436, -4107, 9418, -4086,
9397, -4081, 9392, -4079, 9390, -4078, 9389, -4065,
9376, -4056, 9367, -4041, 9352, -4040, 9351, -4038,
9349, -4035, 9346, -4030, 9341, -4026, 9337, -4020,
9331, -4015, 9326, -4014, 9325, -4010, 9321, -3991,
9302, -3988, 9299, -3984, 9295, -3980, 9291, -3978,
9289, -3977, 9288, -3976, 9287, -3971, 9282, -3970,
9281, -3950, 9261, -3946, 9257, -3938, 9249, -3932,
9243, -3922, 9233, -3920, 9231, -3915, 9226, -3910,
9221, -3909, 9220, -3908, 9219, -3901, 9212, -3900,
9211, -3898, 9209, -3887, 9198, -3885, 9196, -3877,
9188, -3875, 9186, -3869, 9180, -3864, 9175, -3859,
9170, -3854, 9165, -3853, 9164, -3848, 9159, -3839,
9150, -3835, 9146, -3826, 9137, -3821, 9132, -3812,
9123, -3810, 9121, -3807, 9118, -3806, 9117, -3799,
9110, -3797, 9108, -3789, 9100, -3779, 9090, -3777,
9088, -3774, 9085, -3773, 9084, -3769, 9080, -3767,
9078, -3761, 9072, -3751, 9062, -3750, 9061, -3749,
9060, -3748, 9059, -3742, 9053, -3740, 9051, -3731,
9042, -3726, 9037, -3717, 9028, -3715, 9026, -3714,
9025, -3708, 9019, -3704, 9015, -3702, 9013, -3687,
8998, -3677, 8988, -3661, 8972, -3654, 8965, -3653,
8964, -3649, 8960, -3641, 8952, -3635, 8946, -3622,
8933, -3615, 8926, -3610, 8921, -3607, 8918, -3601,
8912, -3597, 8908, -3592, 8903, -3586, 8897, ... , 2594, 2717, 2598, 2713, 2599, 2712, 2603,
2708, 2607, 2704, 2617, 2694, 2619, 2692, 2633,
2678, 2634, 2677, 2643, 2668, 2644, 2667, 2648,
2663, 2650, 2661









share|improve this question











$endgroup$




I wrote a module in Mathematica which finds all possible pairs of integers from a specified list of integers (which can be negative, zero, or positive) which sum to a specified integer m.



The only limiting assumption this algorithm has is that the user only wishes to get the set of all unique sums which sum to m.



Is there a faster algorithm to do this? I've read that making a Hash table is of complexity O(n). Is my code of time O(n)? If it of time O(n), is it a Hash table, or is it something else? If it is not of time O(n), how efficient is it?



FindTwoIntegersWhoseSumIsM[listOfIntegers_,m_]:=Module[

i,distanceFrom1ToMin,negativeFactor,distance,start,finish,(*Integers*)
sortedList,numberLine,temp,finalList,(*Lists*)
execute(*Boolean*)
,
(*There are possible inputted values of m with a give integer set input which
make the execution of this algorithm unnecessary.*)
execute=True;

sortedList=Sort[DeleteDuplicates[listOfIntegers]];

(*Create a continuous list of integers whose smallest and largest entries is equal
to the smallest and largest entries of the inputted list of integers, respectively.*)
(*Let this list be named numberline.*)

(*:::::Construction of numberline BEGINS::::*)

(*If the listOfIntegers only contains negative integers and possibly zero,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]<=0),

(*If m is positive, there is no reason to proceed.*)

If[m>0,execute=False,
(*If m [Equal] 0 then if two or more zeros are in listOfIntegers, they should be outputted to the user.
Therefore, we write m>0 instead of m[GreaterEqual]0 in the conditional above.*)

(*Otherwise, treat it as if all integers were positive with a few considerations.*)
negativeFactor=-1;
sortedList=Reverse[-sortedList];
If[sortedList[[1]]!=0,
numberLine=Range[sortedList[[Length[sortedList]]]]
,
numberLine=Join[0,Range[sortedList[[Length[sortedList]]]]]
]
]
,
negativeFactor=1;

(*Else If the integer set contains negative and positive integers,*)
If[(sortedList[[1]]<0)&&(sortedList[[Length[sortedList]]]>0),
numberLine=
Join[
-Range[Abs[sortedList[[1]]],0,-1](*negative integer subset*)
,
Range[sortedList[[Length[sortedList]]]](*positive integer subset*)
]
,(*Else if the integer set contains only whole numbers,*)
If[(sortedList[[1]]==0)&&(sortedList[[Length[sortedList]]]>0),

(*If the list of integers are all positive and m is negative,
there is no reason to proceed.*)
If[m<0,execute=False,(*Otherwise,*)
numberLine=
Join[
0(*zero*)
,
Range[sortedList[[Length[sortedList]]]](*positive integers*)
]
]
,(*Else if the integer set contains only the natural numbers.*)

(*If the list of integers are all positive and m is negative or zero,
there is no reason to proceed.*)
If[m<=0,execute=False,numberLine=Range[Max[sortedList](*positive integers*)]]
]
]
];

(*:::::Construction of numberline ENDS::::*)
(*Print[numberLine];*)


If[execute==False,finalList=$Failed,
(*Mark all numbers which are in numberline but are not in listOfIntegers with a period.

Sort[] will still sort this list of mixed precision of numbers in ascending order.*)
temp=Sort[Join[Complement[numberLine,sortedList]//N,sortedList]];

(*The main idea of the algorithm is to find the point on numberline to begin selecting two number
combinations which sum to m. m is obviously going to be used when that time comes.

Once that point is selected, integers symmetrically equally distant apart from each other
on both sides of this point (number) in numberline are candidates which sum to m.

To avoid going "out of bounds" of numberline (from either attempting to select a value smaller
than the minimum value of numberline or attempting to select a larger value than the maximum
value of numberline, the following is the maximum distance we can use to obtain ALL possible
two integer combinations which sum to m but of which also prevents us from going "out of bounds".)
*)


(*If the numberline we are about to create had a consistent minimum value of 1
then it would not be offset as it is in general.
The following takes this "offset" into account.*)
distanceFrom1ToMin=Abs[1-Min[sortedList]];


distance=
Min[

distanceFrom1ToMin+Floor[negativeFactor*m/2]
,
Length[temp]-(distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1)

];

start=distanceFrom1ToMin+Floor[negativeFactor*m/2]+1;
finish=distanceFrom1ToMin+Ceiling[negativeFactor*m/2]-1;

(*With the bound distance established, we are ready to begin selecting numbers from numberline.*)
finalList=;
i=1;
While[i<=distance,
finalList=Append[finalList,temp[[start-i]],temp[[finish+i]]];
i++
];

(*It turns out that for even m the first selected integer combination considered is m/2,m/2.*)
If[(Mod[m,2]==0)&&(MemberQ[finalList,negativeFactor*m/2,negativeFactor*m/2]==True),
(*Should there not be two of m/2 in listOfIntegers, we omit this selected combination.*)
If[Length[Flatten[Position[listOfIntegers,negativeFactor*m/2]]]<2,
finalList=Delete[finalList,Position[finalList,negativeFactor*m/2,negativeFactor*m/2][[1]][[1]]]
]
];

(*We selected all possible number combinations in numberline. However, unless listOfIntegers
is all consecutive integers, we need to omit any selected number combination in which either
of the numbers has a "." to the right of it.*)
finalList=negativeFactor*Sort[Select[finalList,Precision[#]==[Infinity]&]]
];
finalList
]


I did the following tests with the code and got these results. (The first number in the time in second it took to do the computation. But you can of course copy the code and do tests yourself.) I omitted most of the results from the last test because it made my post too large, but you will see that it did the computation in 0.209207 seconds.



As the comments in my algorithm (and the algorithm itself suggests), I broke up the number line into negative integers, zero, and the positive integers. I therefore wrote my tests to address all possible situations.




For the positive (non-zero) integer set.



With positive m such that m is larger than what any two number combination in listOfIntegers could possibly sum to.



m = 100; listOfIntegers = RandomSample[Range[20], 6]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

19, 11, 1, 4, 13, 17

0.0371008,


With positive odd m.



m = 215; listOfIntegers = RandomSample[Range[266], 190]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

119, 175, 7, 123, 42, 173, 15, 56, 233, 41, 9, 156, 130, 196, 183,
65, 102, 109, 177, 161, 230, 105, 91, 103, 146, 47, 234, 133, 88, 68,
169, 197, 46, 198, 108, 263, 205, 129, 4, 157, 245, 210, 203, 78,
172, 128, 138, 61, 262, 159, 148, 45, 225, 239, 72, 74, 151, 34, 36,
5, 106, 77, 223, 116, 8, 2, 11, 54, 124, 87, 221, 213, 171, 93, 53,
19, 40, 30, 95, 215, 39, 140, 49, 158, 94, 38, 28, 247, 84, 75, 257,
33, 163, 132, 69, 211, 193, 222, 114, 240, 32, 149, 167, 135, 107,
115, 101, 100, 166, 144, 251, 253, 224, 154, 48, 44, 26, 181, 259,
81, 6, 70, 122, 255, 189, 235, 112, 110, 174, 85, 147, 117, 18, 209,
66, 121, 155, 206, 207, 212, 98, 113, 254, 214, 178, 111, 227, 165,
204, 231, 194, 20, 176, 150, 162, 241, 243, 199, 90, 55, 127, 191,
12, 185, 242, 125, 265, 25, 1, 250, 201, 168, 76, 134, 266, 82, 10,
92, 143, 217, 126, 218, 182, 220, 153, 164, 216, 238, 67, 14

0.136695, 1, 214, 2, 213, 4, 211, 5, 210, 6, 209, 8,
207, 9, 206, 10, 205, 11, 204, 12, 203, 14, 201, 18,
197, 19, 196, 26, 189, 30, 185, 32, 183, 33, 182, 34,
181, 38, 177, 39, 176, 40, 175, 41, 174, 42, 173, 44,
171, 46, 169, 47, 168, 48, 167, 49, 166, 53, 162, 54,
161, 56, 159, 61, 154, 65, 150, 66, 149, 67, 148, 68,
147, 69, 146, 72, 143, 75, 140, 77, 138, 81, 134, 82,
133, 85, 130, 87, 128, 88, 127, 90, 125, 91, 124, 92,
123, 93, 122, 94, 121, 98, 117, 100, 115, 101,
114, 102, 113, 103, 112, 105, 110, 106, 109, 107, 108


With positive even m.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.00998522, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With positive even m such that listOfIntegers contains two of m/2.



m = 22; listOfIntegers = Append[Range[20], 11]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 11

0.00037181, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12, 11, 11


With positive even m such that listOfIntegers contains one m/2.



m = 22; listOfIntegers = Range[20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20

0.000267311, 2, 20, 3, 19, 4, 18, 5, 17, 6, 16, 7,
15, 8, 14, 9, 13, 10, 12


With any negative m.



m = -6; listOfIntegers = Range[26]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25, 26

0.000108231, $Failed



For the positive integer set (including 0).



With an even m.



m = 88; listOfIntegers = RandomSample[Join[0, Range[122]], 39]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

121, 69, 120, 56, 36, 55, 17, 114, 7, 59, 32, 4, 20, 79, 92, 62, 50,
89, 13, 70, 113, 75, 76, 80, 108, 53, 83, 95, 0, 85, 86, 77, 10, 54,
48, 66, 104, 100, 35

0.000505232, 13, 75, 32, 56, 35, 53


With an odd m.



m = 57; listOfIntegers = RandomSample[Join[0, Range[82]], 52]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

62, 18, 26, 0, 67, 34, 55, 52, 35, 78, 10, 68, 46, 44, 38, 23, 77,
76, 58, 51, 75, 63, 53, 42, 54, 27, 56, 71, 12, 17, 2, 37, 31, 72,
49, 50, 32, 16, 47, 19, 4, 20, 81, 25, 61, 14, 80, 82, 59, 33, 70, 39

0.000372743, 2, 55, 4, 53, 10, 47, 18, 39, 19, 38, 20,
37, 23, 34, 25, 32, 26, 31



For the negative integer set (including 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-2, -16, -15, -9, -5, -12, -8, -22, -7, -21, -13, -18, -4, -11, -10,
-19, -6, -17, -20

0.000105898, $Failed


With a negative odd m.



m = -17; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-5, -1, -10, -13, -15, -19, -2, 0, -7, -18, -3, -21, -8, -11, -12,
-22, -17, -16, -20

0.000640987, 0, -17, -1, -16, -2, -15, -5, -12, -7, -10


With a negative even m.



m = -26; listOfIntegers = 
RandomSample[Join[0, -Range[22, 1, -1]], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -16, -11, -14, -17, -13, -1, -9, -15, -20, -18, -4, -21, 0, -8,
-6, -10, -7, -3

0.000329357, -6, -20, -7, -19, -8, -18, -9, -17, -10,
-16, -11, -15



For the negative integer set (excluding 0).



With a positive m.



m = 4; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-20, -7, -16, -21, -11, -13, -5, -2, -6, -19, -1, -12, -18, -14,
-15, -9, -4, -17, -22

0.000102633, $Failed


With a negative odd m.



m = -27; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-18, -17, -22, -13, -1, -11, -19, -8, -16, -6, -21, -12, -20, -3,
-4, -9, -7, -14, -15

0.000242586, -6, -21, -7, -20, -8, -19, -9, -18, -11,
-16, -12, -15, -13, -14


With a negative even m.



m = -26; listOfIntegers = RandomSample[-Range[22, 1, -1], 19]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-19, -10, -20, -9, -21, -14, -5, -1, -17, -4, -18, -22, -8, -6, -13,
-3, -2, -12, -15

0.000286438, -4, -22, -5, -21, -6, -20, -8, -18, -9, -17,
-12, -14



For the complete integer set.



With a positive odd m.



m = 15; listOfIntegers = 
RandomSample[Join[-Range[52, 1, -1], 0, Range[52]], 35]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-30, 19, 42, 38, -25, 6, 48, 5, -8, -27, -11, -47, -37, -12, -3,
-34, 50, 11, 10, 18, 7, -15, 51, -22, -26, -2, 33, -35, 34, 39, 44,
-51, -33, -16, -23

0.000468378, -35, 50, -33, 48, -27, 42, -23, 38, -3,
18, 5, 10


With a negative odd m.



m = -7; listOfIntegers = 
RandomSample[Join[-Range[22, 1, -1], 0, Range[22]], 21]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-1, -16, -11, 10, 17, 1, 0, -5, -22, 8, -7, 15, 21, 11, 18, 14, -4,
7, -13, 4, -9

0.000310697, -22, 15, -11, 4, -7, 0


With a positive even m.



m = 36; listOfIntegers = 
RandomSample[Join[-Range[30, 1, -1], 0, Range[30]], 20]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

25, -9, -8, 8, 5, -10, -24, 13, 9, -16, -23, -14, -22, -29, 26, 12,
19, 16, -30, 18

0.000289237,


With a negative even m.



m = -34; listOfIntegers = 
RandomSample[Join[-Range[100, 1, -1], 0, Range[100]], 50]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

7, 92, 91, 58, -58, 63, -95, 82, 26, 60, 16, 65, 15, 34, 29, 67, -2,
88, 21, -72, -93, 12, 43, 18, -83, -80, -30, -6, 54, -13, -63, 39,
-55, 9, -78, 5, -16, 52, -24, -82, -18, 2, -90, 37, -60, 80, 57, -22,
-26, 72

0.000726359, -63, 29, -60, 26, -55, 21, -18, -16


With m == 0.



m = 0; listOfIntegers = 
RandomSample[Join[-Range[222, 1, -1], 0, Range[222]], 111]
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]
Clear[m, listOfIntegers]

-215, -8, 186, 153, 17, 83, 149, -45, -18, 14, -161, 6, 84, -41,
-59, -130, 34, -24, -142, -95, -70, -60, -152, 90, -43, 12, -196,
-98, -193, -78, -192, 7, -30, 218, -209, -28, -125, 142, 11, 161,
-143, -135, -212, 134, 1, -177, -100, 2, 63, -180, -50, 79, -129,
-91, 126, 57, -140, -200, 38, -182, -107, -25, -46, -179, -113, 88,
148, 28, 184, -158, 190, -9, -36, -5, 169, 221, -204, -210, 44, 45,
-71, 40, 135, 119, -42, 166, 65, 59, -15, -118, 117, -47, -52, 102,
74, -19, 152, 81, 0, 170, -214, 114, -38, 210, -1, -7, -89, -173,
123, 78, -127

0.00179934, -210, 210, -161, 161, -152, 152, -142,
142, -135, 135, -78, 78, -59, 59, -45, 45, -38,
38, -28, 28, -7, 7, -1, 1


With a large m with a large listOfIntegers.



m = 5311; listOfIntegers = 
RandomSample[Join[-Range[9999, 1, -1], 0, Range[9999]], 8888];
AbsoluteTiming[FindTwoIntegersWhoseSumIsM[listOfIntegers, m]]

0.209207, -4680, 9991, -4676, 9987, -4664, 9975, -4650,
9961, -4646, 9957, -4645, 9956, -4636, 9947, -4634,
9945, -4633, 9944, -4630, 9941, -4600, 9911, -4599,
9910, -4594, 9905, -4587, 9898, -4574, 9885, -4573,
9884, -4572, 9883, -4566, 9877, -4562, 9873, -4556,
9867, -4549, 9860, -4538, 9849, -4529, 9840, -4517,
9828, -4514, 9825, -4511, 9822, -4504, 9815, -4502,
9813, -4499, 9810, -4497, 9808, -4490, 9801, -4486,
9797, -4485, 9796, -4483, 9794, -4481, 9792, -4478,
9789, -4475, 9786, -4464, 9775, -4463, 9774, -4458,
9769, -4452, 9763, -4443, 9754, -4431, 9742, -4428,
9739, -4427, 9738, -4420, 9731, -4417, 9728, -4407,
9718, -4405, 9716, -4397, 9708, -4394, 9705, -4393,
9704, -4380, 9691, -4377, 9688, -4369, 9680, -4359,
9670, -4356, 9667, -4354, 9665, -4350, 9661, -4349,
9660, -4346, 9657, -4337, 9648, -4332, 9643, -4331,
9642, -4325, 9636, -4323, 9634, -4314, 9625, -4305,
9616, -4293, 9604, -4283, 9594, -4266, 9577, -4246,
9557, -4241, 9552, -4235, 9546, -4231, 9542, -4227,
9538, -4224, 9535, -4222, 9533, -4220, 9531, -4211,
9522, -4203, 9514, -4202, 9513, -4198, 9509, -4196,
9507, -4193, 9504, -4190, 9501, -4181, 9492, -4176,
9487, -4148, 9459, -4138, 9449, -4137, 9448, -4136,
9447, -4127, 9438, -4125, 9436, -4107, 9418, -4086,
9397, -4081, 9392, -4079, 9390, -4078, 9389, -4065,
9376, -4056, 9367, -4041, 9352, -4040, 9351, -4038,
9349, -4035, 9346, -4030, 9341, -4026, 9337, -4020,
9331, -4015, 9326, -4014, 9325, -4010, 9321, -3991,
9302, -3988, 9299, -3984, 9295, -3980, 9291, -3978,
9289, -3977, 9288, -3976, 9287, -3971, 9282, -3970,
9281, -3950, 9261, -3946, 9257, -3938, 9249, -3932,
9243, -3922, 9233, -3920, 9231, -3915, 9226, -3910,
9221, -3909, 9220, -3908, 9219, -3901, 9212, -3900,
9211, -3898, 9209, -3887, 9198, -3885, 9196, -3877,
9188, -3875, 9186, -3869, 9180, -3864, 9175, -3859,
9170, -3854, 9165, -3853, 9164, -3848, 9159, -3839,
9150, -3835, 9146, -3826, 9137, -3821, 9132, -3812,
9123, -3810, 9121, -3807, 9118, -3806, 9117, -3799,
9110, -3797, 9108, -3789, 9100, -3779, 9090, -3777,
9088, -3774, 9085, -3773, 9084, -3769, 9080, -3767,
9078, -3761, 9072, -3751, 9062, -3750, 9061, -3749,
9060, -3748, 9059, -3742, 9053, -3740, 9051, -3731,
9042, -3726, 9037, -3717, 9028, -3715, 9026, -3714,
9025, -3708, 9019, -3704, 9015, -3702, 9013, -3687,
8998, -3677, 8988, -3661, 8972, -3654, 8965, -3653,
8964, -3649, 8960, -3641, 8952, -3635, 8946, -3622,
8933, -3615, 8926, -3610, 8921, -3607, 8918, -3601,
8912, -3597, 8908, -3592, 8903, -3586, 8897, ... , 2594, 2717, 2598, 2713, 2599, 2712, 2603,
2708, 2607, 2704, 2617, 2694, 2619, 2692, 2633,
2678, 2634, 2677, 2643, 2668, 2644, 2667, 2648,
2663, 2650, 2661






algorithm code-review






share|improve this question















share|improve this question













share|improve this question




share|improve this question








edited Mar 25 at 18:04









Henrik Schumacher

64.3k5 gold badges92 silver badges178 bronze badges




64.3k5 gold badges92 silver badges178 bronze badges










asked Mar 25 at 17:45









Christopher MowlaChristopher Mowla

1186 bronze badges




1186 bronze badges











  • $begingroup$
    The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
    $endgroup$
    – Henrik Schumacher
    Mar 25 at 18:04










  • $begingroup$
    You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
    $endgroup$
    – MikeY
    Mar 25 at 18:37










  • $begingroup$
    According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:58










  • $begingroup$
    Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
    $endgroup$
    – Daniel Lichtblau
    Mar 27 at 23:15
















  • $begingroup$
    The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
    $endgroup$
    – Henrik Schumacher
    Mar 25 at 18:04










  • $begingroup$
    You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
    $endgroup$
    – MikeY
    Mar 25 at 18:37










  • $begingroup$
    According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:58










  • $begingroup$
    Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
    $endgroup$
    – Daniel Lichtblau
    Mar 27 at 23:15















$begingroup$
The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
$endgroup$
– Henrik Schumacher
Mar 25 at 18:04




$begingroup$
The presence of an Append indicates that the complexity of the algorithm is larger than you expect...
$endgroup$
– Henrik Schumacher
Mar 25 at 18:04












$begingroup$
You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
$endgroup$
– MikeY
Mar 25 at 18:37




$begingroup$
You have a Sort call. Use SortBy instead, it is much faster than Sort. But you probably don't need to sort it anyway.
$endgroup$
– MikeY
Mar 25 at 18:37












$begingroup$
According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
$endgroup$
– Christopher Mowla
Mar 25 at 21:58




$begingroup$
According to my knowledge, I did need to use some type of sort for my algorithm. However, I clearly see now (by Roman's post) that my algorithm isn't the most efficient out there. So I guess I'm not worried about it anymore. I wrote this algorithm as part as my coding challenge for a position at Wolfram Research about four months ago. I was just curious if someone could identify what I did or if it is a new way to approach this old classic problem. Thanks, guys!
$endgroup$
– Christopher Mowla
Mar 25 at 21:58












$begingroup$
Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
$endgroup$
– Daniel Lichtblau
Mar 27 at 23:15




$begingroup$
Assuming the hashing and lookup are O(1), an O(n) method is as follows. (1) Hash all values in the list. (2) Iterate over the list, checking for each value k whether m-k was hashed. Can use Sow to record the pair, and Reap to gather all pairs sown.
$endgroup$
– Daniel Lichtblau
Mar 27 at 23:15










2 Answers
2






active

oldest

votes


















16












$begingroup$

I think



IntegerPartitions[m, 2, listOfIntegers]


does exactly what you want, and seems pretty efficient.






share|improve this answer









$endgroup$












  • $begingroup$
    Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:52






  • 2




    $begingroup$
    @ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
    $endgroup$
    – Andrew Cheong
    Mar 26 at 5:28



















2












$begingroup$

Your algorithm certainly isn't $O(n)$, just because sort has a theoretical lower bound of $O(n log(n))$. I don't know what exactly you're trying to do in your algorithm, but the problem is a classic one. One solving strategy is:



  1. Sort the list of numbers

  2. For each number $a_i$ in the list, conduct a binary search for the number $a_j= m - a_i$. If there is such a number $a_j$ for $i neq j$, output the tuple $(a_i,a_j)$.

Put together, this algorithm yields a runtime of $O(nlog(n))$, because binary search runs in $O(log(n))$ and you need to do it $n$ times and sorting is $O(nlog(n))$ anyways.






share|improve this answer









$endgroup$








  • 1




    $begingroup$
    You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
    $endgroup$
    – TonyK
    Mar 26 at 14:21










  • $begingroup$
    @Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16










  • $begingroup$
    If you would like me to make a short video explanation of the algorithm (with an example), I can.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16













Your Answer








StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "387"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);

else
createEditor();

);

function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);



);













draft saved

draft discarded


















StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmathematica.stackexchange.com%2fquestions%2f193945%2fthe-most-efficient-algorithm-to-find-all-possible-integer-pairs-which-sum-to-a-g%23new-answer', 'question_page');

);

Post as a guest















Required, but never shown

























2 Answers
2






active

oldest

votes








2 Answers
2






active

oldest

votes









active

oldest

votes






active

oldest

votes









16












$begingroup$

I think



IntegerPartitions[m, 2, listOfIntegers]


does exactly what you want, and seems pretty efficient.






share|improve this answer









$endgroup$












  • $begingroup$
    Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:52






  • 2




    $begingroup$
    @ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
    $endgroup$
    – Andrew Cheong
    Mar 26 at 5:28
















16












$begingroup$

I think



IntegerPartitions[m, 2, listOfIntegers]


does exactly what you want, and seems pretty efficient.






share|improve this answer









$endgroup$












  • $begingroup$
    Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:52






  • 2




    $begingroup$
    @ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
    $endgroup$
    – Andrew Cheong
    Mar 26 at 5:28














16












16








16





$begingroup$

I think



IntegerPartitions[m, 2, listOfIntegers]


does exactly what you want, and seems pretty efficient.






share|improve this answer









$endgroup$



I think



IntegerPartitions[m, 2, listOfIntegers]


does exactly what you want, and seems pretty efficient.







share|improve this answer












share|improve this answer



share|improve this answer










answered Mar 25 at 18:10









RomanRoman

12.6k1 gold badge19 silver badges50 bronze badges




12.6k1 gold badge19 silver badges50 bronze badges











  • $begingroup$
    Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:52






  • 2




    $begingroup$
    @ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
    $endgroup$
    – Andrew Cheong
    Mar 26 at 5:28

















  • $begingroup$
    Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
    $endgroup$
    – Christopher Mowla
    Mar 25 at 21:52






  • 2




    $begingroup$
    @ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
    $endgroup$
    – Andrew Cheong
    Mar 26 at 5:28
















$begingroup$
Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
$endgroup$
– Christopher Mowla
Mar 25 at 21:52




$begingroup$
Thanks! I have used IntegerPartitions[] for some Rubik's cube theory in the past (cycle types), but I didn't know that it can be used to select partitions from a custom list. I ran it and my algorithm on a larger data set than the largest listed on here, and my algorithm took 31 seconds, whereas IntegerPartitions took 16 seconds. Impressive. I originally wrote my algorithm as part of a coding interview at Wolfram Research, but they didn't hire me for the position. I guess I see why now. LOL.
$endgroup$
– Christopher Mowla
Mar 25 at 21:52




2




2




$begingroup$
@ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
$endgroup$
– Andrew Cheong
Mar 26 at 5:28





$begingroup$
@ChristopherMowla - But also, native functions use pre-compiled code, so have an advantage over your solution. See blog.wolfram.com/2011/12/07/…, especially the second point, if you'd like to optimize your code a bit for better comparison.
$endgroup$
– Andrew Cheong
Mar 26 at 5:28














2












$begingroup$

Your algorithm certainly isn't $O(n)$, just because sort has a theoretical lower bound of $O(n log(n))$. I don't know what exactly you're trying to do in your algorithm, but the problem is a classic one. One solving strategy is:



  1. Sort the list of numbers

  2. For each number $a_i$ in the list, conduct a binary search for the number $a_j= m - a_i$. If there is such a number $a_j$ for $i neq j$, output the tuple $(a_i,a_j)$.

Put together, this algorithm yields a runtime of $O(nlog(n))$, because binary search runs in $O(log(n))$ and you need to do it $n$ times and sorting is $O(nlog(n))$ anyways.






share|improve this answer









$endgroup$








  • 1




    $begingroup$
    You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
    $endgroup$
    – TonyK
    Mar 26 at 14:21










  • $begingroup$
    @Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16










  • $begingroup$
    If you would like me to make a short video explanation of the algorithm (with an example), I can.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16















2












$begingroup$

Your algorithm certainly isn't $O(n)$, just because sort has a theoretical lower bound of $O(n log(n))$. I don't know what exactly you're trying to do in your algorithm, but the problem is a classic one. One solving strategy is:



  1. Sort the list of numbers

  2. For each number $a_i$ in the list, conduct a binary search for the number $a_j= m - a_i$. If there is such a number $a_j$ for $i neq j$, output the tuple $(a_i,a_j)$.

Put together, this algorithm yields a runtime of $O(nlog(n))$, because binary search runs in $O(log(n))$ and you need to do it $n$ times and sorting is $O(nlog(n))$ anyways.






share|improve this answer









$endgroup$








  • 1




    $begingroup$
    You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
    $endgroup$
    – TonyK
    Mar 26 at 14:21










  • $begingroup$
    @Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16










  • $begingroup$
    If you would like me to make a short video explanation of the algorithm (with an example), I can.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16













2












2








2





$begingroup$

Your algorithm certainly isn't $O(n)$, just because sort has a theoretical lower bound of $O(n log(n))$. I don't know what exactly you're trying to do in your algorithm, but the problem is a classic one. One solving strategy is:



  1. Sort the list of numbers

  2. For each number $a_i$ in the list, conduct a binary search for the number $a_j= m - a_i$. If there is such a number $a_j$ for $i neq j$, output the tuple $(a_i,a_j)$.

Put together, this algorithm yields a runtime of $O(nlog(n))$, because binary search runs in $O(log(n))$ and you need to do it $n$ times and sorting is $O(nlog(n))$ anyways.






share|improve this answer









$endgroup$



Your algorithm certainly isn't $O(n)$, just because sort has a theoretical lower bound of $O(n log(n))$. I don't know what exactly you're trying to do in your algorithm, but the problem is a classic one. One solving strategy is:



  1. Sort the list of numbers

  2. For each number $a_i$ in the list, conduct a binary search for the number $a_j= m - a_i$. If there is such a number $a_j$ for $i neq j$, output the tuple $(a_i,a_j)$.

Put together, this algorithm yields a runtime of $O(nlog(n))$, because binary search runs in $O(log(n))$ and you need to do it $n$ times and sorting is $O(nlog(n))$ anyways.







share|improve this answer












share|improve this answer



share|improve this answer










answered Mar 26 at 10:07









Simon ErniSimon Erni

211 bronze badge




211 bronze badge







  • 1




    $begingroup$
    You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
    $endgroup$
    – TonyK
    Mar 26 at 14:21










  • $begingroup$
    @Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16










  • $begingroup$
    If you would like me to make a short video explanation of the algorithm (with an example), I can.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16












  • 1




    $begingroup$
    You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
    $endgroup$
    – TonyK
    Mar 26 at 14:21










  • $begingroup$
    @Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16










  • $begingroup$
    If you would like me to make a short video explanation of the algorithm (with an example), I can.
    $endgroup$
    – Christopher Mowla
    Mar 26 at 17:16







1




1




$begingroup$
You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
$endgroup$
– TonyK
Mar 26 at 14:21




$begingroup$
You don't need to conduct any binary searches! Just make two simultaneous passes through the sorted list, one in increasing order and one in decreasing order. If the sum of the two scanned elements is less than $m$, increase the lower one; else decrease the higher one. Repeat until they meet in the middle. This is $O(n)$. (Of course the total complexity is still $O(nlog n)$ because of the sorting stage.)
$endgroup$
– TonyK
Mar 26 at 14:21












$begingroup$
@Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
$endgroup$
– Christopher Mowla
Mar 26 at 17:16




$begingroup$
@Simon, even though my code may look long, the majority of it is comments explaining the algorithm. If you are interested in knowing what I DID (not what I'm "trying to do") with my algorithm, start reading the comments from the line which states "The main idea of the algorithm is to find the point on numberline to begin selecting two number combinations which sum to m". All lines of code previous to that are for creating the list named numberLine and determining values of some booleans to end the program if there is some trivial reason (based off of m) why there is no need to execute.
$endgroup$
– Christopher Mowla
Mar 26 at 17:16












$begingroup$
If you would like me to make a short video explanation of the algorithm (with an example), I can.
$endgroup$
– Christopher Mowla
Mar 26 at 17:16




$begingroup$
If you would like me to make a short video explanation of the algorithm (with an example), I can.
$endgroup$
– Christopher Mowla
Mar 26 at 17:16

















draft saved

draft discarded
















































Thanks for contributing an answer to Mathematica Stack Exchange!


  • Please be sure to answer the question. Provide details and share your research!

But avoid


  • Asking for help, clarification, or responding to other answers.

  • Making statements based on opinion; back them up with references or personal experience.

Use MathJax to format equations. MathJax reference.


To learn more, see our tips on writing great answers.




draft saved


draft discarded














StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmathematica.stackexchange.com%2fquestions%2f193945%2fthe-most-efficient-algorithm-to-find-all-possible-integer-pairs-which-sum-to-a-g%23new-answer', 'question_page');

);

Post as a guest















Required, but never shown





















































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown

































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown







Popular posts from this blog

Kamusi Yaliyomo Aina za kamusi | Muundo wa kamusi | Faida za kamusi | Dhima ya picha katika kamusi | Marejeo | Tazama pia | Viungo vya nje | UrambazajiKuhusu kamusiGo-SwahiliWiki-KamusiKamusi ya Kiswahili na Kiingerezakuihariri na kuongeza habari

Swift 4 - func physicsWorld not invoked on collision? The Next CEO of Stack OverflowHow to call Objective-C code from Swift#ifdef replacement in the Swift language@selector() in Swift?#pragma mark in Swift?Swift for loop: for index, element in array?dispatch_after - GCD in Swift?Swift Beta performance: sorting arraysSplit a String into an array in Swift?The use of Swift 3 @objc inference in Swift 4 mode is deprecated?How to optimize UITableViewCell, because my UITableView lags

Access current req object everywhere in Node.js ExpressWhy are global variables considered bad practice? (node.js)Using req & res across functionsHow do I get the path to the current script with Node.js?What is Node.js' Connect, Express and “middleware”?Node.js w/ express error handling in callbackHow to access the GET parameters after “?” in Express?Modify Node.js req object parametersAccess “app” variable inside of ExpressJS/ConnectJS middleware?Node.js Express app - request objectAngular Http Module considered middleware?Session variables in ExpressJSAdd properties to the req object in expressjs with Typescript