# Collatz Conjecture

 Unsolved problem in mathematics : Does the Collatz sequence eventually reach 1 for all positive integer initial values?

The Collatz conjecture is aconjecture inmathematics named afterLothar Collatz, who first proposed it in 1937. The conjecture is also known as the 3 n  + 1 conjecture , the Ulam conjecture (afterStanisław Ulam), Kakutani’s problem (afterShizuo Kakutani), the Thwaites conjecture (after Sir Bryan Thwaites), Hasse’s algorithm (afterHelmut Hasse), or the Syracuse problem ;the sequence of numbers involved is referred to as the hailstone sequence or hailstone numbers (because the values are usually subject to multiple descents and ascents likehailstones in a cloud),or as wondrous numbers .

The conjecture can be summarized as follows. Take anypositive integer n . If n is even, divide it by 2 to get n / 2. If n is odd, multiply it by 3 and add 1 to obtain 3 n + 1. Repeat the process (which has been called "Half Or Triple Plus One", or HOTPO ) indefinitely. The conjecture is that no matter what number you start with, you will always eventually reach 1.

Paul Erdős said about the Collatz conjecture: "Mathematics may not be ready for such problems."He also offered \$500 for its solution.

## Contents

• 1 Statement of the problem
• 5 Supporting arguments
• 5.1 Experimental evidence
• 5.2 A probabilistic heuristic
• 6 Other formulations of the conjecture
• 6.2 As an abstract machine that computes in base two
• 6.3 As a parity sequence
• 7 Extensions to larger domains
• 7.1 Iterating on all integers
• 7.2 Iterating with odd denominators or 2-adic integers
• 7.3 Iterating on real or complex numbers
• 7.3.1 Collatz fractal
• 8.2 Modular restrictions
• 10 Undecidable generalizations
• 11 The Ultimate Challenge: the 3 x +1 problem

## Statement of the problem [ edit ]

Numbers from 1 to 9999 and their corresponding total stopping time

Histogram of total stopping times for the numbers 1 to 100 million. Total stopping time is on the x axis, frequency on the y axis. Note that for all positive integers the histogram would be a completely different, exponentially-growing sequence (see)

Consider the following operation on an arbitrarypositive integer:

• If the number is even, divide it by two.
• If the number is odd, triple it and add one.

Inmodular arithmetic notation, define thefunction f as follows:

Now, form a sequence by performing this operation repeatedly, beginning with any positive integer, and taking the result at each step as the input at the next.

In notation:

(that is: is the value of applied to recursively times; ).

The Collatz conjecture is: This process will eventually reach the number 1, regardless of which positive integer is chosen initially.

That smallest i such that a i = 1 is called the total stopping time of n .The conjecture asserts that every n has a well-defined total stopping time. If, for some n , such an i doesn’t exist, we say that n has infinite total stopping time and the conjecture is false.

If the conjecture is false, it can only be because there is some starting number which gives rise to a sequence that does not contain 1. Such a sequence might enter a repeating cycle that excludes 1, or increase without bound. No such sequence has been found.

## Examples [ edit ]

For instance, starting with n = 6, one gets the sequence 6, 3, 10, 5, 16, 8, 4, 2, 1.

n = 19, for example, takes longer to reach 1: 19, 58, 29, 88, 44, 22, 11, 34, 17, 52, 26, 13, 40, 20, 10, 5, 16, 8, 4, 2, 1.

The sequence for n = 27, listed and graphed below, takes 111 steps (41 steps through odd numbers ), climbing to 9232 before descending to 1.

27 , 82, 41 , 124, 62, 31 , 94, 47 , 142, 71 , 214, 107 , 322, 161 , 484, 242, 121 , 364, 182, 91 , 274, 137 , 412, 206, 103 , 310, 155 , 466, 233 , 700, 350, 175 , 526, 263 , 790, 395 , 1186, 593 , 1780, 890, 445 , 1336, 668, 334, 167 , 502, 251 , 754, 377 , 1132, 566, 283 , 850, 425 , 1276, 638, 319 , 958, 479 , 1438, 719 , 2158, 1079 , 3238, 1619 , 4858, 2429 , 7288, 3644, 1822, 911 , 2734, 1367 , 4102, 2051 , 6154, 3077 , 9232 , 4616, 2308, 1154, 577 , 1732, 866, 433 , 1300, 650, 325 , 976, 488, 244, 122, 61 , 184, 92, 46, 23 , 70, 35 , 106, 53 , 160, 80, 40, 20, 10, 5 , 16, 8, 4, 2, 1 (sequence inOEIS)

Numbers with a total stopping time longer than any smaller starting value form a sequence beginning with:

1, 2, 3, 6, 7, 9, 18, 25, 27, 54, 73, 97, 129, 171, 231, 313, 327, 649, 703, 871, 1161, 2223, 2463, 2919, 3711, 6171, … (sequence inOEIS).

The starting values whose maximum trajectory point is greater than that of any smaller starting value are as follows:

1, 2, 3, 7, 15, 27, 255, 447, 639, 703, 1819, 4255, 4591, 9663, 20895, 26623, 31911, 60975, 77671, 113383, 138367, 159487, 270271, 665215, 704511, … (sequence inOEIS)

Number of steps for n to reach 1 are

0, 1, 7, 2, 5, 8, 16, 3, 19, 6, 14, 9, 9, 17, 17, 4, 12, 20, 20, 7, 7, 15, 15, 10, 23, 10, 111, 18, 18, 18, 106, 5, 26, 13, 13, 21, 21, 21, 34, 8, 109, 8, 29, 16, 16, 16, 104, 11, 24, 24, … (sequence inOEIS)

The longest progression for any initial starting number less than 100 million is 63,728,127, which has 949 steps. For starting numbers less than 1 billion it is 670,617,279, with 986 steps, and for numbers less than 10 billion it is 9,780,657,631, with 1132 steps.

Thepowers of two converge to one quickly because is halved times to reach one, and is never increased, but forMersenne number M n , they need to increase n times and usually need more steps to reach 1.

## Visualizations [ edit ]

• Directed graph showing the orbits of small numbers under the Collatz map. The Collatz conjecture is equivalent to the statement that all paths eventually lead to 1.

• Directed graph showing the orbits of the first 1000 numbers.

• The x axis represents starting number, the y axis represents the highest number reached during the chain to 1.

## Cycles [ edit ]

Any counterexample to the Collatz conjecture would have to consist either of an infinite divergent trajectory or a cycle different from the trivial (4; 2; 1) cycle. Thus, if one could prove that neither of these types of counterexample could exist, then all positive integers would have a trajectory that reaches the trivial cycle. Such a strong result is not known, but certain types of cycles have been ruled out.

The type of a cycle may be defined with reference to the "shortcut" definition of the Collatz map, for odd n and for even n . A cycle is a sequence where , , and so on, up to in a closed loop. For this shortcut definition, the only known cycle is (1; 2). Although 4 is part of the single known cycle for the original Collatz map, it is not part of the cycle for the shortcut map.

A k -cycle is a cycle that can be partitioned into 2 k contiguous subsequences: k increasing sequences of odd numbers alternating with k decreasing sequences of even numbers. For instance, if the cycle consists of a single increasing sequence of odd numbers followed by a decreasing sequence of even numbers, it is called a 1-cycle .

Steiner (1977) proved that there is no 1-cycle other than the trivial (1;2).Simons (2004) used Steiner’s method to prove that there is no 2-cycle.Simons & de Weger (2003) extended this proof up to 68-cycles: there is no k -cycle up to k = 68.Beyond 68, this method gives upper bounds for the elements in such a cycle: for example, if there is a 75-cycle, then at least one element of the cycle is less than 2385×2 50 .Therefore, as exhaustive computer searches continue, larger cycles may be ruled out. To state the argument more intuitively: we need not look for cycles that have at most 68 trajectories, where each trajectory consists of consecutive ups followed by consecutive downs. Seefor an idea of how one might find an upper bound for the elements of a cycle.

## Supporting arguments [ edit ]

Although the conjecture has not been proven, most mathematicians who have looked into the problem think the conjecture is true because experimental evidence and heuristic arguments support it.

### Experimental evidence [ edit ]

The conjecture has been checked by computer for all starting values up to 2 60 .All initial values tested so far eventually end in the repeating cycle (4; 2; 1), which has only three terms. From this lower bound on the starting value, a lower bound can also be obtained for the number of terms a repeating cycle other than (4; 2; 1) must have.When this relationship was established in 1981, the formula gave a lower bound of 35,400 terms.

This computer evidence is not a proof that the conjecture is true. As shown in the cases of thePólya conjecture, theMertens conjecture and theSkewes’ number, sometimes a conjecture’s onlycounterexamples are found when using very large numbers.

### A probabilistic heuristic [ edit ]

If one considers only the odd numbers in the sequence generated by the Collatz process, then each odd number is on average 3/4 of the previous one.(More precisely, the geometric mean of the ratios of outcomes is 3/4.) This yields a heuristic argument that every Hailstone sequence should decrease in the long run, although this is not evidence against other cycles, only against divergence. The argument is not a proof because it assumes that Hailstone sequences are assembled from uncorrelated probabilistic events. (It does rigorously establish that the2-adic extension of the Collatz process has two division steps for every multiplication step for almost all 2-adic starting values.)

And even if the probabilistic reasoning were rigorous, this would still imply only that the conjecture isalmost surely true for any given integer, which does not necessarily imply that it is true for all integers.

### Rigorous bounds [ edit ]

Although it is not known rigorously whether all positive numbers eventually reach one according to the Collatz iteration, it is known that many numbers do so. In particular, Krasikov and Lagarias showed that the number of integers in the interval [1, x ] that eventually reach one is at least proportional to x 0.84 .

## Other formulations of the conjecture [ edit ]

### In reverse [ edit ]

The first 21 levels of the Collatzgraph generated in bottom-up fashion. The graph includes all numbers with an orbit length of 21 or less.

There is another approach to prove the conjecture, which considers the bottom-up method of growing the so-called Collatz graph . The Collatz graph is agraph defined by the inverserelation

So, instead of proving that all positive integers eventually lead to 1, we can try to prove that 1 leads to all positive integers. For any integer n , n ≡ 1 (mod 2)iff 3 n + 1 ≡ 4 (mod 6). Equivalently, ( n − 1)/3 ≡ 1 (mod 2) iff n ≡ 4 (mod 6). Conjecturally, this inverse relation forms atree except for the 1–2–4 loop (the inverse of the 4–2–1 loop of the unaltered function f defined in the Statement of the problem section of this article).

When the relation 3 n + 1 of the function f is replaced by the common substitute "shortcut" relation (3 n + 1)/2, the Collatz graph is defined by the inverse relation,

For any integer n , n ≡ 1 (mod 2) iff (3 n + 1)/2 ≡ 2 (mod 3). Equivalently, (2 n − 1)/3 ≡ 1 (mod 2) iff n ≡ 2 (mod 3). Conjecturally, this inverse relation forms a tree except for a 1–2 loop (the inverse of the 1–2 loop of the function f(n) revised as indicated above).

Alternately, replace the 3 n + 1 with n’ / H(n’) where n’ = 3 n + 1 and H(n’) is the highest power of 2 that divides n’ (with noremainder). The resulting function f maps fromodd numbers to odd numbers. Now suppose that for some odd number n , applying this operation k times yields the number 1 (that is, ). Then inbinary, the number n can be written as the concatenation ofstrings w k w k-1 … w 1 where each w h is a finite and contiguous extract from the representation of 1 / 3 h .The representation of n therefore holds therepetends of 1 / 3 h , where each repetend is optionally rotated and then replicated up to a finite number of bits. It is only in binary that this occurs.Conjecturally, every binary string s that ends with a ‘1’ can be reached by a representation of this form (where we may add or delete leading ‘0’s to s ).

### As an abstract machine that computes in base two [ edit ]

Repeated applications of the Collatz function can be represented as anabstract machine that handlesstrings ofbits. The machine will perform the following three steps on any odd number until only one "1" remains:

1. Append 1 to the (right) end of the number in binary (giving 2 n  + 1);
2. Add this to the original number by binary addition (giving 2 n  + 1 +  n = 3 n  + 1);
3. Remove all trailing "0"s (i.e. repeatedly divide by two until the result is odd).

This prescription is plainly equivalent to computing a Hailstone sequence in base two.

Example [ edit ]

The starting number 7 is written in base two as 111. The resulting Hailstone sequence is:

### As a parity sequence [ edit ]

For this section, consider the Collatz function in the slightly modified form

This can be done because when n is odd, 3 n + 1 is always even.

If P(…) is the parity of a number, that is P(2 n ) = 0 and P(2 n + 1) = 1, then we can define the Hailstone parity sequence (or parity vector) for a number n as p i = P( a i ), where a 0 = n , and a i +1 = f ( a i ).

What operation is performed (3n + 1)/2 or n/2 depends on the parity. The parity sequence is the same as the sequence of operations.

Using this form for f ( n ), it can be shown that the parity sequences for two numbers m and n will agree in the first k terms if and only if m and n are equivalent modulo 2 k . This implies that every number is uniquely identified by its parity sequence, and moreover that if there are multiple Hailstone cycles, then their corresponding parity cycles must be different.

Applying the f function k times to the number a ·2 k b will give the result a ·3 c d , where d is the result of applying the f function k times to b , and c is how many odd numbers were encountered during that sequence.

### As a tag system [ edit ]

For the Collatz function in the form

Hailstone sequences can be computed by the extremely simple2-tag system with production rules abc , ba , caaa . In this system, the positive integer n is represented by a string of n a , and iteration of the tag operation halts on any word of length less than 2. (Adapted from De Mol.)

The Collatz conjecture equivalently states that this tag system, with an arbitrary finite string of a’ s as the initial word, eventually halts (see Tag system#Example: Computation of Collatz sequences for a worked example).

## Extensions to larger domains [ edit ]

### Iterating on all integers [ edit ]

An obvious extension is to include all integers, not just positive integers. Leaving aside the trivial cycle 0 → 0, there are a total of 4 known non-trivial cycles, which all nonzero integers seem to eventually fall into under iteration of f . These cycles are listed here, starting with the well-known cycle for positive  n :

Odd values are listed in large bold. Each cycle is listed with its member of least absolute value (which is always odd) first.

Cycle Odd-value cycle length Full cycle length
1 → 4 → 2 → 1 1 3
−1 → −2 → −1 1 2
−5 → −14 → −7 → −20 → −10 → −5 2 5
−17 → −50 → −25 → −74 → −37 → −110 → −55 → −164 → −82 → −41 → −122 → −61 → −182 → −91 → −272 → −136 → −68 → −34 → −17 7 18

The generalized Collatz conjecture is the assertion that every integer, under iteration by f , eventually falls into one of the four non-trivial cycles above, or is the trivial cycle 0 → 0.

### Iterating with odd denominators or 2-adic integers [ edit ]

The standard Collatz map can be extended to (positive or negative) rational numbers which have odd denominators when written in lowest terms. The number is taken to be odd or even according to whether its numerator is odd or even. A closely related fact is that the Collatz map extends to the ring of2-adic integers, which contains the ring of rationals with odd denominators as a subring.

The parity sequences as defined above are no longer unique for fractions. However, it can be shown that any possible parity cycle is the parity sequence for exactly one fraction: if a cycle has length n and includes odd numbers exactly m times at indices k 0 , …, k m −1 , then the unique fraction which generates that parity cycle is

( 1 )

For example, the parity cycle (1 0 1 1 0 0 1) has length 7 and has 4 odd numbers at indices 0, 2, 3, and 6. The unique fraction which generates that parity cycle is

the complete cycle being: 151/47 → 250/47 → 125/47 → 211/47 → 340/47 → 170/47 → 85/47 → 151/47

Although the cyclic permutations of the original parity sequence are unique fractions, the cycle is not unique, each permutation’s fraction being the next number in the loop cycle:

(0 1 1 0 0 1 1) →
(1 1 0 0 1 1 0) →
(1 0 0 1 1 0 1) →
(0 0 1 1 0 1 1) →
(0 1 1 0 1 1 0) →
(1 1 0 1 1 0 0) →

Also, for uniqueness, the parity sequence should be "prime", i.e., not partitionable into identical sub-sequences. For example, parity sequence (1 1 0 0 1 1 0 0) can be partitioned into two identical sub-sequences (1 1 0 0)(1 1 0 0). Calculating the 8-element sequence fraction gives

(1 1 0 0 1 1 0 0) →

But when reduced to lowest terms {5/7}, it is the same as that of the 4-element sub-sequence

(1 1 0 0) →

And this is because the 8-element parity sequence actually represents two circuits of the loop cycle defined by the 4-element parity sequence.

In this context, the Collatz conjecture is equivalent to saying that (0 1) is the only cycle which is generated by positive whole numbers (i.e. 1 and 2).

Cycle bounds [ edit ]

also gives a rough idea about how one can prove that cycles of certain lengths do not exist. For a hypothetical cycle of length n , the numerator is bounded above by 3 n – 2 n (this corresponds to a cycle of all odd numbers). A lower bound for the denominator can be obtained by letting n / m be an optimal rational approximation to log(3)/log(2). Together these give an upper bound for the unique fraction that generates a cycle of length n . If this upper bound is smaller than thelargest number for which the conjecture has been verified to hold, then a cycle of length n is impossible.

### Iterating on real or complex numbers [ edit ]

Cobweb plot of the orbit 10-5-8-4-2-1-2-1-2-1-etc. in the real extension of the Collatz map (optimized by replacing "3 n + 1" with "(3 n + 1)/2")

The Collatz map can be viewed as the restriction to the integers of the smooth real and complex map

which simplifies to

If the standard Collatz map defined above is optimized by replacing the relation 3 n + 1 with the common substitute "shortcut" relation (3 n + 1)/2, it can be viewed as the restriction to the integers of the smooth real and complex map

which simplifies to .

Collatz fractal [ edit ]

Iterating the above optimized map in the complex plane produces the Collatzfractal.

The point of view of iteration on the real line was investigated by Chamberland (1996),and on the complex plane by Letherman, Schleicher, and Wood (1999).

Collatz mapfractal in a neighbourhood of the real line

## Optimizations [ edit ]

### Time-space tradeoff [ edit ]

The sectionabove gives a way to speed up simulation of the sequence. To jump ahead k steps on each iteration (using the f function from that section), break up the current number into two parts, b (the k least significant bits, interpreted as an integer), and a (the rest of the bits as an integer). The result of jumping ahead k steps can be found as:

f k ( a 2 k + b ) = a 3 c (b) + d (b).

The c and d arrays are precalculated for all possible k -bit numbers b , where d (b) is the result of applying the f function k times to b , and c (b) is the number of odd numbers encountered on the way.For example, if k=5, you can jump ahead 5 steps on each iteration by separating out the 5 least significant bits of a number and using:

c (0..31) = {0,3,2,2,2,2,2,4,1,4,1,3,2,2,3,4,1,2,3,3,1,1,3,3,2,3,2,4,3,3,4,5}
d (0..31) = {0,2,1,1,2,2,2,20,1,26,1,10,4,4,13,40,2,5,17,17,2,2,20,20,8,22,8,71,26,26,80,242}.

This requires 2 k precomputation and storage to speed up the resulting calculation by a factor of k , aspace-time tradeoff.

### Modular restrictions [ edit ]

For the special purpose of searching for a counterexample to the Collatz conjecture, this precomputation leads to an even more important acceleration, used by Tomás Oliveira e Silva in his computational confirmations of the Collatz conjecture up to large values of n . If, for some given b and k , the inequality

f k ( a 2 k b ) = a 3 c (b) + d (b) < a 2 k b

holds for all a , then the first counterexample, if it exists, cannot be b modulo 2 k .For instance, the first counterexample must be odd because f (2 n ) = n , smaller than 2 n ; and it must be 3 mod 4 because f 2 (4 n + 1) = 3 n + 1, smaller than 4 n + 1. For each starting value a which is not a counterexample to the Collatz conjecture, there is a k for which such an inequality holds, so checking the Collatz conjecture for one starting value is as good as checking an entire congruence class. As k increases, the search only needs to check those residues b that are not eliminated by lower values of  k . Only an exponentially small fraction of the residues survive.For example, the only surviving residues mod 32 are 7, 15, 27, and 31.

## Syracuse function [ edit ]

If is an odd integer, then is even, so with odd and . The Syracuse function is the function from the set of odd integers into itself, for which (sequence inOEIS).

Some properties of the Syracuse function are:

The Collatz conjecture is equivalent to the statement that, for all in , there exists an integer such that .

## Undecidable generalizations [ edit ]

In 1972,J. H. Conway proved that a natural generalization of the Collatz problem is algorithmicallyundecidable.

Specifically, he considered functions of the form

where are rational numbers which are so chosen that is always integral.

The standard Collatz function is given by , , , , . Conway proved that the problem:

Given g and n , does the sequence of iterates reach 1?

is undecidable, by representing thehalting problem in this way. Closer to the Collatz problem is the following universally quantified problem:

Given g does the sequence of iterates reach 1, for all n>0 ?

Modifying the condition in this way can make a problem either harder or easier to solve (intuitively, it is harder to justify a positive answer but might be easier to justify a negative one). Kurtz and Simonproved that the above problem is, in fact, undecidable and even higher in the arithmetical hierarchy , specifically -complete. This hardness result holds even if one restricts the class of functions g by fixing the modulus P to 6480.

## The Ultimate Challenge: the 3 x +1 problem [ edit ]

This volume,edited byJeffrey Lagarias and published by the American Mathematical Society , is a compendium of information on the Collatz conjecture, methods of approaching it and generalizations. It includes two survey papers by the editor and five by other authors, concerning the history of the problem, generalizations, statistical approaches and results from the theory of computation . It also includes reprints of early papers on the subject (including an entry by Lothar Collatz himself).

## References [ edit ]

1. Maddux, Cleborne D.; Johnson, D. Lamont (1997). Logo: A Retrospective . New York: Haworth Press. p. 160.ISBN  0-7890-0374-0 . The problem is also known by several other names, including: Ulam’s conjecture, the Hailstone problem, the Syracuse problem, Kakutani’s problem, Hasse’s algorithm, and the Collatz problem.
2. Jeffrey C. Lagarias (1985). The 3x + 1 problem and its generalizations. The American Mathematical Monthly 92(1): 3-23.
3. According to Lagarias (1985),p. 4, the name "Syracuse problem" was proposed by Hasse in the 1950s, during a visit toSyracuse University.
4. Pickover, Clifford A. (2001). Wonders of Numbers . Oxford: Oxford University Press. pp. 116–118.ISBN  0-19-513342-0 .
5. "Hailstone Number" . MathWorld . Wolfram Research.
6. Hofstadter, Douglas R. (1979). Gödel, Escher, Bach . New York: Basic Books. pp. 400–402.ISBN  0-465-02685-0 .
7. Friendly, Michael (1988). Advanced Logo: A Language for Learning . Hillsdale, New Jersey, USA: Lawrence Erlbaum Associates.ISBN  0-89859-933-4 .
8. Guy, Richard K. (2004). Unsolved problems in number theory (3rd ed.).Springer-Verlag. "E17: Permutation Sequences".ISBN  0-387-20860-7 . Zbl   1058.11001 . Cf pp. 336–337 .
9. Guy (2004),p. 330
10. R. K. Guy: Don’t try to solve these problems, Amer. Math. Monthly, 90 (1983), 35–41. By this Erdos means that there aren’t powerful tools for manipulating such objects.
11. Leavens, Gary T.; Vermeulen, Mike (December 1992). "3x+1 Search Programs". Computers & Mathematics with Applications 24 (11): 79–99.doi: 10.1016/0898-1221(92)90034-F .
12. Roosendaal, Eric. "3x+1 Delay Records" . Retrieved 27 November 2011 . (Note: "Delay records" are total stopping time records.)
13. Simons, J.; de Weger, B.; "Theoretical and computational bounds for m -cycles of the 3 n  + 1 problem" , Acta Arithmetica (on-line version 1.0, November 18, 2003), 2005.
14. Steiner, R. P.; "A theorem on the syracuse problem", Proceedings of the 7th Manitoba Conference on Numerical Mathematics , pages 553–559, 1977.
15. Silva, Tomás Oliveira e Silva. "Computational verification of the 3x+1 conjecture" . Retrieved 13 May 2015 .
16. Garner, Lynn E. "On The Collatz 3n + 1 Algorithm" (PDF) . Retrieved 27 March 2015 .
17. Lagarias (1985),section " A heuristic argument" .
18. Krasikov, Ilia;Lagarias, Jeffrey C. (2003). "Bounds for the 3 x  + 1 problem using difference inequalities". Acta Arithmetica 109 (3): 237–258.doi: 10.4064/aa109-3-4 . MR   1980260 .
19. Colussi, Livio (9 September 2011). "The convergence classes of Collatz function". Theoretical Computer Science 412 (39): 5409–5419.doi: 10.1016/j.tcs.2011.05.056 .
20. Hew, Patrick Chisan (7 March 2016). "Working in binary protects the repetends of 1/3h: Comment on Colussi’s ‘The convergence classes of Collatz function’". Theoretical Computer Science 618 : 135–141.doi: 10.1016/j.tcs.2015.12.033 .
21. Terras, Riho (1976), "A stopping time problem on the positive integers" (PDF) , Polska Akademia Nauk 30 (3): 241–252,MR  0568274
22. Marc Chamberland. A continuous extension of the 3 x  + 1 problem to the real line. Dynam. Contin. Discrete Impuls Systems 2: 4 (1996), 495–509.
23. Simon Letherman, Dierk Schleicher, and Reg Wood: The (3 n  + 1)-Problem and Holomorphic Dynamics. Experimental Mathematics 8: 3 (1999), 241–252.
24. Scollo, Giuseppe (2007), "Looking for Class Records in the 3x+1 Problem by means of the COMETA Grid Infrastructure", Grid Open Days at the University of Palermo (PDF)
25. Lagarias (1985),Theorem D.
26. Conway, John H. (1972), "Unpredictable Iterations", Proc. 1972 Number Theory Conf., Univ. Colorado, Boulder , pp. 49–52
27. Kurtz, Stuart A.; Simon, Janos (2007). "The Undecidability of the Generalized Collatz Problem". In Cai, J.-Y.; Cooper, S.B.; Zhu, H. Proceedings of the 4th International Conference on Theory and Applications of Models of Computation, TAMC 2007, held in Shanghai, China in May 2007 . pp. 542–553.doi: 10.1007/978-3-540-72504-6_49 . ISBN   3-540-72503-2 . Also available here: http://www.cs.uchicago.edu/~simon/RES/collatz.pdf
28. Ben-Amram, Amir M. (2015), "Mortality of iterated piecewise affine functions over the integers: Decidability and complexity", Computability 1 (1): 19–56,doi: 10.3233/COM-150032
29. Lagarias, Jeffrey C., ed. (2010). The Ultimate Challenge: the 3x+1 problem . American Mathematical Society . ISBN   978-0-8218-4940-8 . Zbl   1253.11003 .