Perfect Totient Number¶

An integer greater than two whose successive Euler totients through one add up exactly to the starting integer.

Version: v1 · 2026-10-25 · History
Domain-specific #: 14428
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Number Theory → Mathematics

Core Idea¶

A perfect totient number is an integer \(n>2\) for which the sum of all its successive Euler-totient values after \(n\), ending with \(1\), equals \(n\).^[1] Let \(φ\) be Euler's totient, and write \(φ^i\) for \(i\) successive applications. If \(\varphi^c(n)=2\), the test is \(n=\sum_{i=1}^{c+1}\varphi^i(n)\). The starting \(n\) is excluded; the final \(\varphi^{c+1}(n)=1\) is included.^[1] Neither equality to a divisor sum nor merely having a totient orbit meets the test.

The equality can be recognized in unlike arithmetic settings. A partition of the integers from \(1\) through \(15\) into gcd classes has sizes \(8,4,2,1\), which the authors notice are exactly the successive totients from \(15\); their sum returns \(15\).^[2] In another setting, a theorem covers every positive power of \(3\); at \(27=3^3\), the independently computed orbit \(18,6,2,1\) also returns its start.^[1]^[2] The partition is a way to discover one case, while the totient orbit and exact equality define both cases.

Structural Signature¶

Sig role-map: eligible integer \(n>2\) → repeated Euler totient → post-start orbit through \(1\) → sum compared exactly with \(n\).

Eligible input. The same integer supplies the starting point of the computation and the target for its sum.^[1] Replacing it with an arbitrary state changes the named number class.
Totient transition. Euler's \(φ\) supplies every next value.^[1] A different recurrence may have its own exact-return numbers, but they are not perfect totient numbers.
Complete post-start orbit. The sum starts at \(\varphi(n)\), follows each iterate, and includes the final \(1\).^[1] A truncated orbit or the addition of \(n\) tests a different equality.
Exact-return decision. Membership holds precisely when that sum equals the initial integer.^[1] A nearby value or an unverified partial calculation does not certify membership.

What It Is Not¶

A perfect number sums its proper positive divisors to the original integer. The shared adjective and equality target do not make its divisor aggregate a totient orbit. A highly totient number concerns how many inputs have a specified totient value. A sparsely totient number compares totient values of larger inputs. Neither takes the successive forward totients of one input and asks for their sum to return it.

The gcd-class construction for \(15\) is not a required ingredient. It led Loomis and colleagues from Gauss's divisor-sum identity to the new equality, but a perfect totient number need not be supplied with that partition.^[2] Nor does the fact that the totient orbit terminates at \(1\) imply that its sum equals the start: \(21\) fails the equality.

Scope of Application¶

The literal habitat is elementary and research number theory: positive integers greater than \(2\), Euler's totient, its iterates, and an exact finite sum.^[1] One can test a particular integer by computing its complete orbit. One can prove an entire family belongs by showing the equality for every eligible member, as the prime-power theorem does for \(3^k\).^[2]

The equality does not range over arbitrary dynamical systems, approximate numerical convergence, or generic forms of “perfection.” At \(15\) the same terms also arise as gcd-class sizes; this combinatorial interpretation explains the discovery but does not extend the definition to arbitrary partitions.^[2]

Clarity¶

The four-role test settles three common ambiguities. “Iterated totients” means successive applications to the last value, not totients of every divisor. “Sum” excludes the starting value and includes \(1\). “Perfect” names exact return to \(n\), not closeness to \(n\). Thus \(21\), despite a valid terminating orbit \(12,4,2,1\), is excluded because the terms total \(19\).^[2] The arithmetic in this negative case follows the standard totient rules stated by Loomis and colleagues; the full \(21\) orbit is our calculation, not a quoted example.^[2]

Manages Complexity¶

A long chain can be reduced to one repeatable test: generate each \(φ\) value, stop at \(1\), sum the generated values, compare once with \(n\). The method separates the membership certificate for a particular input from a family proof. The \(15\) calculation is a complete certificate in the source; the \(3^k\) statement has its own theorem; a list of computed members alone does not prove a whole family.^[2]

This compression also protects the boundary. To check a candidate outside a familiar family, no gcd partition or prior theorem is necessary; the orbit and sum suffice. To claim infinitely many or all members of a parametric form, however, one needs an argument covering the parameter, not a handful of successful orbits.

Abstract Reasoning¶

Given an integer \(n>2\), compute \(\varphi(n)\), then continue until \(1\), accumulating every post-start value. Equality proves membership; inequality proves nonmembership. For \(27\), \(\varphi(27)=18\), \(\varphi(18)=6\), \(\varphi(6)=2\), and \(\varphi(2)=1\), so \(18+6+2+1=27\). These four numerical steps are an independently reproduced check of the source-proved \(3^k\) family, rather than a quoted orbit from the paper.^[1]^[2]

The same reasoning can expose an overbroad claim. A condition on the prime in a product such as \(3p\) cannot be treated as sufficient merely because it is necessary. The original literature gives a stronger conditional characterization involving a second perfect totient number.^[1] Dropping that qualification changes the theorem.

Knowledge Transfer¶

Within number theory, the role map transfers from a worked gcd-class discovery to an algebraically proved family and to direct tests of exceptional candidates. The invariant is the post-start \(φ\) orbit and exact equality, not the way a candidate was found.^[2]^[1]

One might analogize the exact-return test to other iterated maps, but changing \(φ\) would create another class. Such an analogy can suggest a research question; it does not make this named integer class a cross-domain Prime.

Examples¶

Canonical: the source-worked gcd classes of 15¶

For \(n=15\), Gauss's gcd classes partition \(1,\ldots,15\) into four sets of sizes \(8,4,2,1\).^[2] Loomis and colleagues explicitly identify those sizes, in order, with \(\varphi(15)\), \(\varphi(8)\), \(\varphi(4)\), and \(\varphi(2)\).^[2] Their total \(8+4+2+1=15\) is a complete perfect-totient calculation, found through the partition.^[2]

Mapped back: eligible input \(15\) → Euler-totient transition \(15\mapsto8\mapsto4\mapsto2\mapsto1\) → full post-start orbit \(8,4,2,1\) → sum \(15\) equals input. The gcd classes illuminate the values; the equality is what gives membership.

Applied: the prime-power family at 27¶

Iannucci, Deng and Cohen state that all positive powers of \(3\) are perfect totient numbers; Loomis and colleagues prove the prime-power result in their Theorem 2.^[1]^[2] Instantiate it at \(n=27=3^3\): the calculated orbit \(18,6,2,1\) sums to \(27\).^[2] Here an algebraic family theorem, rather than a gcd-class partition, supplies the general warrant; the displayed orbit checks this member independently.

Mapped back: eligible input \(27\) → the same \(φ\) transition → complete post-start orbit \(18,6,2,1\) → exact sum \(27\). As a negative control, \(21\mapsto12\mapsto4\mapsto2\mapsto1\) yields \(19\), so sharing the process without exact return is insufficient. These \(27\) and \(21\) orbit lists are reproduced calculations from elementary totient rules.^[2]

Structural Tensions¶

No universal structural tradeoff is established by these sources between computation and proof, or between a gcd-class discovery and the orbit definition. They are different ways to find or warrant cases. The practical diagnostic is: Does the claim concern this one computed integer, or every integer in a stated family? A complete orbit settles the first; the second needs a theorem whose hypotheses cover the family. A source-listed exceptional input is still an individual membership claim; it does not establish a general family theorem.

Structural–Framed Character¶

The entry is strongly structural within number theory: membership is an exact relation between a specified integer, its totient orbit and a sum. It has no evaluative weight beyond that classification; “perfect” here does not praise an integer. Human practice chooses which numbers to investigate and whether to search or prove a family, but does not alter the equality for a fixed input. Historical naming and publication supply provenance, while no institution confers membership. The words “iteration,” “sum” and “return” travel to other fields, but the full name applies literally only when Euler's totient on integers fills the transition role. A new eligible integer satisfying the equality is recognition of this class; applying the name to a non-totient recurrence is import of an analogy. Its character: a domain-specific arithmetic class with a reusable test but fixed mathematical carrier.

Structural Core vs. Domain Accent¶

The skeletal relation is an iterated transformation whose subsequent values aggregate to its starting value. That pattern could prompt a future Prime question, but the reviewed cases all retain positive integers, Euler's \(φ\), a terminal \(1\), and exact addition. Those are constitutive conditions, not decorative accents. A general “returning orbit sum” might recur on unlike carriers; this evidence does not establish that recurrence. The named perfect totient number therefore remains domain-specific, whether the candidate is found by a partition, a family theorem, or a direct computation.

This entry presupposes Euler's totient function.

The direct parent is Euler's Totient Function under a strict composition/presupposes edge. Without that specific function, the post-start orbit and its sum are undefined as perfect-totient membership; the function itself applies independently to numbers such as \(21\), which fail the equality. The class is not a kind of function, and the function is not an internal part of one integer. Perfect Number, Highly Totient Number, and Sparsely Totient Number are catalog neighbors distinguished by their divisor sum, preimage count, and later-input comparison tests, respectively. No direct Prime edge follows merely from the generic idea of satisfying an equality.

Relationships to Other Abstractions¶

Current abstraction Perfect Totient Number Domain-specific

Parents (1) — outgoing typed relations

Perfect Totient Number presupposes Euler's totient function Domain-specific

Perfect-totient membership requires Euler's totient to generate every post-start value; the function can exist without this number class.

Hierarchy paths (5) — routes to 3 parentless roots

Perfect Totient Number → presupposes Euler's totient function → is a kind of Cardinality → presupposes Bijectivity → is a kind of Function (Mapping)

Show alternative paths (4)

Neighborhood in Abstraction Space¶

Perfect Totient Number sits in a sparse region of the domain-specific corpus (81^st percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2769 abstractions)

Nearest neighbors

Average Order of an Arithmetic Function — 0.83
Perfect digit-to-digit invariant — 0.83
Beck–Chevalley Condition — 0.83
Floyd's Cycle-Finding Algorithm — 0.82
Cumulative Hierarchy — 0.81

Computed from structural-signature embeddings · 2026-10-25

Not to Be Confused With¶

A perfect number. It sums proper divisors; this entry sums successive totient values.
A totient value or inverse-totient count. The former is one transition value, and the latter counts preimages of a value. Neither is the complete exact-return test.
A gcd partition. It explains one worked \(15\) example but is not necessary for membership.

References¶

[1] Douglas E. Iannucci, Deng Moujie, and Graeme L. Cohen, “On Perfect Totient Numbers,” Journal of Integer Sequences 6 (2003), Article 03.4.5, abstract, §1 and Table 1 (PDF pp. 1–2). registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l

[2] Paul Loomis, Michael Plytage, and John Polhill, “Summing Up the Euler Phi Function,” The College Mathematics Journal 39, no. 1 (2008), 34–42, especially pp. 34–37, 39. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p