Almost and quasi Leinster groups

Since P_i is an even perfect number, using 1.1, equation (1.2) becomes:

where 2^r-1 is a Mersenne prime r\geq 2.
We assume, by contradiction, that k is even, i.e. k=2\ell, \ell\in\mathbb{N}^*. Equation (Proof) becomes:

Numbers p^\ell-1 and p^\ell+1 are even consecutive numbers. It follows that r-1\geq 2\iff r\geq 3.
Since 2^r-1 is prime, it divides only one of the factors in the left-hand side (either p^\ell-1, or p^\ell+1). Since they are both even, it follows that 2^{r+1}-2 is a divisor of one of the factors. Because 2^{r+1}-2>2^{r-2}, and p^\ell+1>p^\ell-1, we get p^\ell-1=2^{r-2} and p^\ell+1=2^{r+1}-2. Subtracting these, it follows that 2^{r+1}-4=2^{r-2}, which yields a contradiction, hence k is odd.
Now, we prove that k can actually be just 1.
Assume by contradiction, that k>1. In this case, equation (Proof) becomes:

Since p^k is odd, it follows that p is odd. Together with the fact that k is odd, we get that p^{k-1}+\dots+p+1 is odd. Since p-1 is odd, it follows that 2^{r-1}|(p-1). Also, because 2^r-1 is prime, it divides just one of the factors in the right-hand side. Both the factors in the right hand side are greater than 1; it follows that 2^{r-1}=p-1 and 2^r-1=p^{k-1}+\dots+p+1. We get that: 2p-3=p^{k-1}+\dots+p+1, which is equivalent to p^{k-1}+\dots+p-2p=-4. Since p is odd, this has no solutions.
This concludes the proof.

The direct ((1)\Rightarrow(2)) is Proposition 3.2 from ([Medts]). The converse follows from (1.4).

The converse statement follows from (1.4). We prove the direct.

Since G is nilpotent, it can be written as the direct product of its Sylow p-subgroups:

where |G_i|=p_i^{n_i}, for all i\in\{1,2,\dots,k\}. Since the function D is multiplicative, it follows that:

Thus, we get

We assume, by contradiction, that G_1 is not cyclic. It follows that n_1\geq 2. In addition, the final theorem from [Miler] states that the number of subgroups of index p_1 is congruent to 1+p_1\pmod{p_1^2}; in particular, it is greater or equal to 1+p_1. These subgroups are normal, so D(G)\geq 1+p_1^{\alpha_1}+(p_1+1)p_1^{\alpha_1-1}. It follows that:

Since G is quasi-Leinster and \delta is multiplicative, it follows that

a contradiction. Thus G_1 is cyclic. Analogously, all G_i, i=2,\dots,k, are cyclic and therefore G is cyclic.

According to [Medts, Theorem 5.7], the value of D(ZM(m,n,r)) is given by:

Assume that m is prime, and let d be the order of r modulo m. According to [Medts, Corollary 5.9.], equation (Proof) becomes:

Suppose further that n=d. It follows that:

If we assume that ZM(m,n,r) is quasi-Leinster, it follows from Lemma 3.2 that:

which, using (Proof), implies D(n)=2n, i.e. n is perfect. On the other hand, it follows from (3.1), that

since m is prime, it follows that the order of r modulo m is m-1. Since we have assumed that the order of r is n, m-1=n, or, equivalently m=n+1, which has solutions, according to Example 1.3. Any r\in\{2,\dots,m-1\} satisfies the conditions.

Since \mathbb{F}_q^* is cyclic, it follows from Proposition 4.1 that:

Then

We obtain that:

1. \operatorname{Aff}(\mathbb{F}_q) is Leinster if and only if 1+qD(q-1)=2q(q-1). This yields q|1, which is absurd.

2. \operatorname{Aff}(\mathbb{F}_q) is quasi-Leinster if and only if 1+qD(q-1)=2q(q-1)-1. It follows that q|2, thus q=2, i.e. \operatorname{Aff}(\mathbb{F}_q)\cong C_2.

3. \operatorname{Aff}(\mathbb{F}_q) is almost Leinster if and only if 1+qD(q-1)=2q(q-1)+1. This implies D(q-1)=2(q-1), i.e. q-1 is perfect. According to Corollary 1.2, q=p.

If we assume that D_{2n} is almost/quasi-/Leinster, we get

We compute the value for D(D_{2n}), distinguishing two cases based on the parity of n:

• n odd. Using (5.1) from Lemma 5.1, together with (Proof), we get

  2\cdot 2n+a=2n+D(n), a\in\{-1,0,1\},

  which is equivalent to D(n)=2n\pm a, i.e. n being almost/quasi-/perfect.

• n even. Again, using (Proof) and Lemma 5.1, we get: 4n+D(n)=4n+ a, a\in\{-1,0,1\}, which has no solution.

If \operatorname{Dic}_{4n} is almost/quasi-/Leinster, we get

Using lemma 6.1, we compute D(\operatorname{Dic}_{4n}) based on the parity of n.

• n even. In this case, it follows from (6.1), that D(\operatorname{Dic}_{4n})=D(2n)+8n. Together with (Proof), we get D(2n)+8n=8n + a, a\in\{-1,0,1\}, which has no solution.

• n odd. In this case D(\operatorname{Dic}_{4n})=D(2n)+4n. It follows that D(2n)+4n=8n+a, a\in\{-1,0,1\}, i.e.:

  D(2n)=4n + a, a\in\{-1,0,1\}.

  1. If D(2n)=4n-1, then 2n is almost perfect.

  2. If D(2n)=4n, then 2n is perfect.

  3. If D(2n)=4n+1, then 2n is quasi-perfect.