We study the Marcinkiewicz–Zygmund property for finite differences. A difference D(f,h)=∑k=1nakf(bkh) is said to have the MZ-property if, for every measurable function f, D(f,h)=o(h)andf(h)=o(1) imply f(h)=o(h). In this paper, we characterize all three-node differences D(f,h)=af(qh)+bf(rh)+cf(sh) having the MZ-property. In particular, for irrational differences with distinct nodes, we obtain explicit necessary and sufficient conditions in terms of the coefficients a,b,c and the nodes q,r,s. As an application, our results resolve an open problem concerning a second-order Riemann difference.
We consider three-term functional differences of the form
D(f,h)=af(qh)+bf(rh)+cf(sh),(1)
where h is a real variable and f is an arbitrary complex-valued function defined for real numbers in a neighborhood of 0. Each such difference D has fixed parameters. These include nonzero coefficients, a, b, c, which may be real or complex, and distinct nodes, q, r, s, which are nonzero real numbers. We consider three possible limit properties of f as h→0:
A. D(f,h)=o(h)
B. f(h)=o(1)
C. f(h)=o(h)
Property C trivially implies both A and B, for any choice of parameters. If A and B together imply C then the difference D(f,h) is said to have the MZ-property.
A case requiring special consideration is the following.
Definition 1. A difference (1) is called rational if there are three integers i,j,k, not all zero, such that ilog∣q∣+jlog∣r∣+klog∣s∣=i+j+k=0. Otherwise, it is called irrational.
A characterization of rational differences with the MZ-property is contained in the results of [6] and will be discussed below. In this paper we characterize the irrational three-term differences that have the MZ-property. Our main result is the following:
Theorem 2. Let D(f,h)=af(qh)+bf(rh)+cf(sh) be an irrational difference with distinct nodes, q, r, s, where ∣q∣≤∣r∣≤∣s∣. Then D has the MZ-property if and only if at least one of the following conditions holds
∣a∣+∣b∣≤∣c∣
∣br∣+∣cs∣<∣aq∣
∣a∣+∣c∣≤∣b∣ and ∣cs∣+∣aq∣<∣br∣
Since the MZ-property is invariant under scaling either the coefficients {a,b,c} or the nodes {q,r,s}, we will often rewrite the difference in standard form, with q=c=1, 1≤∣r∣≤∣s∣, and
D(f,h)=af(h)+bf(rh)+f(sh).(2)
Then the definition of rational takes on the more familiar form:
Definition 3. A difference (2) is called rational if and only if log∣s∣log∣r∣∈Q.
Theorem 2 can then be re-written in a less symmetrical, but equivalent form:
Theorem 2 (Alternate version). Let D(f,h)=af(h)+bf(rh)+f(sh) be an irrational difference with distinct nodes, 1,r,s where 1≤∣r∣≤∣s∣. Then D has the MZ-property if and only if at least one of the following conditions holds
∣a∣+∣b∣≤1
∣br∣+∣s∣<∣a∣
∣a∣+1≤∣b∣ and ∣s∣+∣a∣<∣br∣.
Real differences
The original motivation for the MZ-property was to study smoothness properties of real functions. However, as shown in [6], this property cannot be properly studied without also including complex numbers. Hence, we allow complex numbers for the coefficients, a,b,c, as well as the output of the arbitrary function f. In this section we will justify the use of complex numbers for the case that we are only interested in real numbers.
Definition 4. A difference (2) is called real if the coefficients, a,b,c are all real numbers.
Definition 5. A difference (2) has the real MZ-property if A and B imply C for all real-valued functions.
Theorem 6. A real difference (2) has the MZ-property if and only if it has the real MZ-property.
Proof. First, suppose that a real difference, D, has the MZ-property. Then properties A and B together imply C for any function f:R→C. Trivially, A and B together imply C for any function f:R→R.
Next, suppose that a real difference fails the MZ-property. Then there is a function f:R→C such that D(f,h)=o(h), f(h)=o(1), but f(h) is not o(h). Let r(h)=Re(f), and c(h)=Im(f), so that r:R→R and c:R→R. Since f=o(1), we have both r=o(1) and c=o(1). Since D(f,h)=o(h), we have Re(D(f,h))=o(h) and Im(D(f,h))=o(h). But since D is a real difference, Re(D(f,h))=D(Re(f),h)=D(r,h), and Im(D(f,h))=D(Im(f),h)=D(c,h).
Finally, if r(h) and c(h) were both o(h), then we would have f(h)=o(h). By choice of f, this does not happen. So, at least one of r(h),c(h) witnesses that D does not have the real MZ-property.
A consequence of this is that Theorem 2 also characterizes the real MZ-property for real differences. For the remainder of this paper, we will only discuss the complex variant.
Background
When an ordinary derivative is not available, one may consider generalized derivatives such as the symmetric derivative
fs′(c)=h→0lim2hf(c+h)−f(c−h).
The symmetric derivative is a generalization of ordinary differentiation since if f′(c) exists then so does fs′(c) and they are equal. Furthermore, it is a proper generalization since
f(h)=∣h∣
is symmetrically differentiable at 0, but not ordinarily differentiable there.
More generally, derivatives of the form
D′f(c)=h→0limh∑k=1nakf(c+bkh).(3)
can be considered, where the coefficients ai are real numbers.
To generalize the ordinary derivative, we need the derivative of a constant to be 0 and the derivative of the identity function to be 1. That is, we need
k=1∑nak=0andk=1∑nakbk=1.(4)
It is not hard to see that these two conditions guarantee that D′f(c) is a generalization of the ordinary derivative. That is, under these two conditions, whenever f′(c) exists, so does D′f(c), and D′f(c)=f′(c). For this reason, derivatives that satisfy these two conditions are called Riemann derivatives.
A Riemann derivative, however, might not yield anything new; it may be equivalent to ordinary differentiation. For example, let
S′f(c)=h→0limh2f(c+h)+f(c−h)−3f(c).
If S′f(c) exists, then f′(c) exists, and f′(c)=S′f(c). So, unlike the symmetric derivative S′ is not a proper generalization of the ordinary derivative.
It is natural to wonder, therefore, which Riemann derivatives are equivalent to the ordinary derivative and which are proper generalizations. Ash, Catoiu, and Csörnyei [1] gave a complete answer to this question. Specifically, a derivative (3) is equivalent to the ordinary derivative if and only if it can be written in the form
where aq=0 and br=1. The derivative S′ above is obtained from R′ by setting a=3 and q=r=b=1.
This raises the question: What if we add continuity as an additional assumption? That is, when does
D′f(c)=kandh→0limf(c+h)=f(c)(5)
imply
f′(c)=k?
To study this question, we assume, without loss of generality, that c=0 and subtract an affine function so that f(c)=k=0. This leads us to the following definition, named after Marcinkiewicz and Zygmund since they are believed to be the first to consider this property (see [7]).
Definition 7. A finite difference
D(f,h)=k=1∑nakf(bkh),
is said to have the MZ-property if and only if for any real function f defined in a neighborhood of 0,
D(f,h)=o(h)andf(h)=o(1)⟹f(h)=o(h).(6)
Note that although the MZ-property was motivated by the study of Riemann derivatives, it may have intrinsic and independent interest, so we make no tacit assumptions on the properties (4) in this paper. Also note, as discussed above, that we have changed this definition to include complex coefficients and functions f:R→C. Finally for simplicity, we have ruled out the possibility that bk=0. This is without loss of generality, since otherwise, D(f,h)=o(h) and f(h)=o(1) imply that f(bkh)=f(0)=0.
Remark 8. In their study of m-th Riemann derivatives2, Marcinkiewicz and Zygmund needed, in place of (6), the implication
D(f,h)=o(hm)andf(h)=o(hm−1)⟹f(h)=o(hm).
Here m is a positive integer, and we call this the MZ-property of order m. However, the MZ-property of order m reduces to the MZ-property defined above. Indeed, define
D1(g,h)=k=1∑nakbkm−1g(bkh).
Then D has the MZ-property of order m if and only if D1 has the MZ-property. To see this, suppose first that D fails the MZ-property of order m for some f, and set
g(h)=hm−1f(h).
Then g(h)=o(1), while
D(f,h)=hm−1D1(g,h).
Thus D1(g,h)=o(h), whereas g(h)=o(h), so D1 fails the MZ-property. Conversely, if D1 fails the MZ-property for some g, then setting
f(h)=hm−1g(h)
shows that D fails the MZ-property of order m.
The MZ-property has been studied for higher order derivatives and for various number of terms, (See [2], [3], and [4].) though a complete characterization has not yet appeared. The primary result of this study solves the case of geometric nodes, defined as follows.
Dmf(c)=h→0limhmD(fc,h),fc(x)=f(c+x).
Definition 9. Let λ be a real number with ∣λ∣>1. Then the finite difference
D(f,h)=k=0∑n(akf(λkh)+ckf(−λkh))
is said to have geometric nodes with ratio λ. It has two characteristic polynomials,
As an example, suppose D(f,h)=5f(−2h)+3f(−h)−5f(0)−7f(2h)+f(23h). Then λ=2, the characteristic polynomials are P+(z)=z3−2z+3, and P−(z)=z3−12z−3.
In the case of geometric nodes, the MZ-property has been completely characterized in terms of the roots of the characteristic polynomials.
Theorem 10. (Fejzić [6]) A difference with geometric nodes has the MZ-property if and only if neither of its characteristic polynomials has a complex root z in the annulus 1<∣z∣≤∣λ∣.
In the course of that investigation, it was necessary to allow for complex coefficients, explaining why we retain this convention, even if our main interest is real-valued functions.
Any two-term difference can be written with geometric nodes after scaling. Therefore, Theorem 10 completely characterizes the two-term case:
Corollary 11. (Fejzić [6]) Let D(f,h)=af(h)+f(qh) be a two-term difference with 1≤∣q∣.
Case 1: If 1=∣q∣, i.e. q=−1 then D(f,h) has the MZ-property if and only if a=±1.
Case 2: If 1<∣q∣ then D(f,h) has the MZ-property if and only if a is not in the annulus 1<∣a∣≤∣q∣.
Proof. Case 1: This can be deduced from Theorem 10, but is also seen directly. Suppose D(f,h)=af(h)+f(−h)=o(h). Then aD(f,h)−D(f,−h)=(a2−1)f(h)=o(h). If a2=1 then f(h)=o(h). For counterexamples, if a=1 use f(h)=h and if a=−1 use f(h)=∣h∣.
Case 2: Let λ=∣q∣. Then, D(f,h)=af(λ0h)+f(λ1h) or D(f,h)=af(λ0h)+f(−λ1h). In any event, the two characteristic polynomials are a+z and a−z, with roots z=±a. By Theorem 10, the MZ-property fails if and only if 1<∣a∣≤∣q∣. □
For rational three-term differences in standard form (2), if
log∣s∣log∣r∣=jk,
for integers k≥0 and j≥1 then letting λ=∣s∣1/j the difference can be written with geometric nodes as D(f,h)=af(h)+bf(sgn(r)λkh)+f(sgn(s)λjh), where ∣λ∣>1. Therefore, Theorem 10 applies, giving the following corollary.
Corollary 12. (Fejzić [6]) Let D(f,h)=af(h)+bf(rh)+f(sh) be a rational difference with 1≤∣r∣≤∣s∣, and let j>0, k≥0 be integers such that log∣s∣log∣r∣=jk. Then D has the MZ-property if and only if the trinomials zj+bzk+a and sgn(s)zj+sgn(r)bzk+a have no roots in the annulus 1<∣z∣≤∣s∣1/j.
The next examples illustrate important differences between the rational and irrational cases. Examples 13 and 14 illustrate that the MZ-property in the rational case can depend on the plus and minus signs attached to coefficients. Examples 13 and 15 illustrate that it can also depend on plus and minus signs attached to the nodes. This is in contrast to the irrational case, Theorem 2, where the MZ-property only depends on absolute values.
Example 13. The difference D(f,h)=f(8h)+4f(2h)+3f(h) has only one distinct trinomial x3+4x+3 whose roots are outside the annulus 1<∣z∣≤2, so it has the MZ-property.
Example 14. The difference D(f,h)=f(8h)−4f(2h)+3f(h) has only one distinct trinomial x3−4x+3. It contains a root, z, where 1<∣z∣≤2, so it does not have the MZ-property. Indeed, here is a specific counterexample. Let f(h)=zn if h=2n and n∈Z. Let f(h)=0 elsewhere.
Example 15. The difference D(f,h)=f(8h)+4f(−2h)+3f(h) has two distinct trinomials, x3+4x+3 and x3−4x+3 and since z=213−1 is a root of the second trinomial with 1<∣z∣≤2, it fails MZ-property. Indeed, one can verify that f(h)=±zn for h=±2nn∈Z and 0 elsewhere is a counterexample.
Accordingly, the main focus of the present paper is the case of irrational three-term differences. Together, Theorem 2 and Corollary 12 completely characterize the MZ-property for three-term differences. But there is a big difference in these results. Theorem 2 solves the irrational case with a direct computation on the coefficients and nodes. Corollary 12 solves the rational case, using the location of the complex roots of polynomial equations. We do not know of a direct characterization for the rational case. In particular, Theorem 2 does not hold for the rational case. This is seen in Example 13 where we have a=3, b=4, r=2, s=8. None of the three conditions of Theorem 2 hold, and yet D(f,h) has the MZ-property. Nevertheless, one direction of Theorem 2 does hold in the rational case.
Theorem 16. Let D(f,h)=af(h)+bf(rh)+f(sh) be a rational difference with distinct nodes, 1,r,s where 1≤∣r∣≤∣s∣. Then D has the MZ-property if at least one of the following conditions holds
∣a∣+∣b∣≤1
∣br∣+∣s∣<∣a∣
∣a∣+1≤∣b∣ and ∣s∣+∣a∣<∣br∣.
We end this section with an application of Theorem 2 to an open problem posed in [5]. Consider the second-order Riemann difference
D(f,h)=f(5h)−2f(2h)+f(−h).
We show that
D(f,h)=o(h2)andf(h)=o(h)
imply
f(h)=o(h2).
By Remark 8, it suffices to show that the difference
D1(g,h)=5g(5h)−4g(2h)−g(−h)
has the MZ-property. Indeed, using the notation of Theorem 2, the coefficients a=−1, b=−4, and c=5 satisfy the first condition of the theorem,
∣a∣+∣b∣≤∣c∣.
The main goal now, is to establish Theorems 2 and 16. The rest of this paper is organized as follows. In Section 3 we establish Theorems 2 and 16 for the edge cases r=−1 and r=−s. This is a mere convenience, allowing us to assume strict inequalities 1<∣r∣<∣s∣ for the remainder of the paper. In Sections 4, 5, and 6 we prove the MZ-property for the three respective conditions stated in Theorem 2 and Theorem 16. In Section 7 we show that if none of the three conditions hold, and D(f,h) is irrational, then the MZ-property fails.
The edge cases r=−1 and r=−s
In this section we consider the difference
D(f,h)=af(h)+bf(rh)+f(sh)
where 1,r,s are distinct, 1≤∣r∣≤∣s∣, and one of the inequalities in not strict. That is, either r=−1 or r=−s.
The cases r=−1 and r=−s do not occur in the irrational case. Indeed, if r=−1 then log∣s∣log∣r∣=0 and if r=−s then log∣s∣log∣r∣=1. In both cases, D(f,h) is rational. The goal of this section, then, is to establish Theorem 16 for these two cases.
Let us first consider r=−1, so that
D(f,h)=af(h)+bf(−h)+f(sh).(8)
with ∣s∣>1.
Corollary 17. A difference (8) has the MZ-property if and only if neither z=∣a+b∣ nor z=∣a−b∣ is in the interval 1<∣z∣≤∣s∣.
Proof. Apply Corollary 12, with k=0 and j=1. The two characteristic polynomials are z+(b+a) and sgn(s)z+(a−b). The absolute values of the respective roots are ∣z∣=∣a+b∣ and ∣z∣=∣a−b∣. The result immediately follows from Corollary 12.
The following corollary is Theorem 16 when r=−1.
Corollary 18.If any of the following conditions hold
∣a∣+∣b∣≤1
∣b∣+∣s∣<∣a∣
∣a∣+1≤∣b∣ and ∣s∣+∣a∣<∣b∣,
then the difference (8) has the MZ-property.
Proof. Note that
max{∣a∣−∣b∣,∣b∣−∣a∣}≤max{∣a−b∣,∣a+b∣}≤∣a∣+∣b∣.
So, if ∣a∣+∣b∣≤1 then ∣a+b∣≤1 and ∣a−b∣≤1. If ∣b∣+∣s∣<∣a∣, then ∣s∣<∣a∣−∣b∣, so ∣s∣<∣a+b∣ and ∣s∣<∣a−b∣. Finally, if ∣s∣+∣a∣<∣b∣ then ∣s∣<∣b∣−∣a∣, so ∣s∣<∣a+b∣ and ∣s∣<∣a−b∣. In all three cases, by Corollary 17, D(f,h) has the MZ-property. □
Next, consider r=−s, so that
D(f,h)=af(h)+bf(−sh)+f(sh).(9)
with ∣s∣>1.
Corollary 19. A difference (9) has the MZ-property if and only if ∣a∣ is not in either of the intervals:
(∣1+b∣,∣s∣∣1+b∣]
(∣1−b∣,∣s∣∣1−b∣]
Proof. Apply Corollary 12, with k=1 and j=1. The two characteristic polynomials are
P+(z)=(1+b)z+aandP−(z)=(sgn(s)+sgn(−s)b)z+a.
Since the parameters are nonzero, if b=−1 then P+ has no root. Otherwise, it has a single root with absolute value ∣z∣=∣1+b∣∣a∣. In other words, P+ has no root, z, in the annulus 1<∣z∣≤∣s∣ if and only if ∣a∣ is not in the interval (∣1+b∣,∣s∣∣1+b∣]. Similarly, P− has no root with
1<∣z∣≤∣s∣ if and only if ∣a∣ is not in the interval (∣1−b∣,∣s∣∣1−b∣]. The result immediately follows by Corollary 12.
The following corollary is Theorem 16 when r=−s.
Corollary 20.If any of the following conditions hold
∣a∣+∣b∣≤1
∣bs∣+∣s∣<∣a∣
∣a∣+1≤∣b∣ and ∣s∣+∣a∣<∣bs∣,
then the difference (9) has the MZ-property.
Proof. If ∣a∣+∣b∣≤1 then ∣a∣≤1−∣b∣≤min{∣1+b∣,∣1−b∣}. If ∣bs∣+∣s∣<∣a∣ then ∣s∣max{∣1+b∣,∣1−b∣}<∣a∣. If ∣a∣+1≤∣b∣ then ∣a∣≤∣b∣−1≤min{∣1+b∣,∣1−b∣}. In all three cases, the MZ-property follows by Corollary 19.
It remains to show Theorems 2 and 16 for the case 1<∣r∣<∣s∣.
The case ∣a∣+∣b∣≤1
In this section we consider
D(f,h)=af(h)+bf(rh)+f(sh)
where 1<∣r∣<∣s∣ and ∣a∣+∣b∣≤1.
In the rational case, suppose log∣s∣log∣r∣=jk for some positive integers, j,k. Then, this result follows from Corollary 12. In fact, no polynomial P(z)=±zj±bzk±a has a root with ∣z∣>1. To see this, let ∣z∣>1. Then,
Therefore, P(z) has no zeros with ∣z∣>1. By Corollary 12, D(f,h) has the MZ-property.
The proof below works directly and does not require any rational relation between log∣r∣ and log∣s∣.
Theorem 21.Let 1<∣r∣<∣s∣, and let a,b∈C satisfy
∣a∣+∣b∣≤1.
Then
Df(h)=af(h)+bf(rh)+f(sh)
has the MZ-property
Proof. Assume Df(h)=o(h) and f(h)=o(1) and fix ε>0. Then, there exists δ>0 such that for all t with 0<∣t∣<δ,
∣af(t)+bf(rt)+f(st)∣≤ε∣t∣<ε∣st∣.
Then,
∣f(st)∣<∣a∣∣f(t)∣+∣b∣∣f(rt)∣+ε∣st∣.(10)
It follows that either
∣f(t)∣>∣f(st)∣−ε∣st∣,or∣f(rt)∣>∣f(st)∣−ε∣st∣.
Indeed, otherwise since ∣a∣+∣b∣≤1, we would get a contradiction to (10). Since also st<δ, either
f(st)>∣f(t)∣−ε∣t∣,orf(srt)>∣f(t)∣−ε∣t∣(11)
Now fix h with ∣h∣<δ. Repeatedly applying 11 define a sequence hm recursively by
h0=h,
and choosing
hm=s1hm−1orhm=srhm−1
so that
∣f(hm)∣>∣f(hm−1)∣−ε∣hm−1∣.(12)
In either case, since 1<∣r∣<∣s∣,
0<∣hm∣≤sr∣hm−1∣.(13)
where 0<sr<1. Therefore, ∣hm∣ strictly decreases to 0, justifying the continued use of (11)
Let c be the sum of the geometric series,
c=j=0∑∞srj.
Iterating (13) gives
0<∣hj∣≤srj∣h∣
for each non-negative integer j and then iterating (12) gives
∣f(hm)∣>∣f(h)∣−εj=0∑m−1∣hj∣>∣f(h)∣−εc∣h∣.
Thus
∣f(h)∣<∣f(hm)∣+εc∣h∣.
Since hm→0 and f(h)=o(1), we have f(hm)→0. Therefore,
∣f(h)∣≤εc∣h∣.
Because ε>0 is arbitrary, it follows that
f(h)=o(h).
The case ∣br∣+∣s∣<∣a∣
In this section we consider
D(f,h)=af(h)+bf(rh)+f(sh)
where 1<∣r∣<∣s∣. We will show that if ∣br∣+∣s∣<∣a∣ then D has the MZ-property.
In the rational case, suppose
log∣s∣log∣r∣=jk
for some positive integers j,k. Then, this result follows from Corollary 12. In fact, no polynomial P(z)=±zj±bzk±a has a root z with ∣z∣≤∣r∣1/k. To see this, let λ=∣r∣1/k. We have λj=∣s∣ and λk=∣r∣. If z is a zero of P(z) then ∣±zj±bzk∣=∣a∣. But when ∣z∣≤λ,
∣±zj±bzk∣≤λj+∣b∣λk=∣s∣+∣b∣∣r∣<∣a∣.
So, P(z) has no zeros with ∣z∣≤λ and by Corollary 12, D(f,h) has the MZ-property.
The following proof works without assuming any rational relation between log∣s∣ and log∣r∣. Furthermore, instead of f(h)=o(1) it requires only f(h)=O(1).
Theorem 22. Let 1<∣r∣<∣s∣, and ∣br∣+∣s∣<∣a∣, and
Df(h)=af(h)+bf(rh)+f(sh).
If
Df(h)=o(h)andf(h)=O(1)
Then
f(h)=o(h).
The proof of this theorem is based on repeated applications of the following lemma.
Lemma 23. Let
Df(h)=af(h)+bf(rh)+f(sh).
with 1<∣r∣<∣s∣, and ∣br∣+∣s∣<∣a∣. Let
t=log∣s∣log∣r∣andK=k=0∑∞(∣a∣∣br∣+∣s∣)k.
Suppose for some α≥0
Df(h)=o(h)andf(h)=O(hα).
Then for each ϵ>0 there is a δ>0 and L>0 such that whenever ∣h∣<δ,
∣f(h)∣<Kϵ∣h∣+L∣h∣α+t.
We first prove the theorem from the lemma:
Proof. Let 1<∣r∣<∣s∣, and ∣br∣+∣s∣<∣a∣, and
Df(h)=af(h)+bf(rh)+f(sh).
Suppose
Df(h)=o(h)andf(h)=O(1).
We will show that f(h)=o(h).
Since f(h)=O(hα) for α=0. We apply the lemma with ϵ=1 and get positive numbers δ1 and L1 such that whenever ∣h∣<δ1,
∣f(h)∣<K∣h∣+L1∣h∣t.
By shrinking δ1, if necessary, we may assume that ∣h∣<1. Since t<1, ∣h∣<∣h∣t. So, for all ∣h∣<δ1,
∣f(h)∣<(K+L1)∣h∣t.
This implies that f(h)=O(ht).
We apply the lemma a second time with α=t and ϵ=1. We get positive numbers δ2 and L2 such that whenever ∣h∣<δ2,
∣f(h)∣<K∣h∣+L2∣h∣2t.
If 2t≤1 then again, by shrinking δ2 if necessary, we get that for all ∣h∣<δ2,
∣f(h)∣<(K+L2)∣h∣2t,
and therefore, f(h)=O(h2t).
This process continues as long as α+t≤1. With each application of the lemma, the value of α increases by the constant t. Eventually, after a finite number of applications, we have f(h)=O(hα) with α≤1 but α+t>1. At that point, we apply the lemma one final time, but this time we fix some arbitrary ϵ>0. By the lemma, there is a δ>0 and L>0 such that whenever ∣h∣<δ,
∣f(h)∣<Kϵ∣h∣+L∣h∣α+t.
Since this time, α+t>1, the first term dominates as h→0. By shrinking δ if necessary, we can assume ∣h∣ is so small that L∣h∣α+t<Kϵ∣h∣. Then we get
∣f(h)∣<2Kϵ∣h∣.
Since ϵ is arbitrary, it follows that f(h)=o(h). □
We now prove the lemma:
Proof. Let
Df(h)=af(h)+bf(rh)+f(sh).
with 1<∣r∣<∣s∣, and ∣br∣+∣s∣<∣a∣. Let
t=log∣s∣log∣r∣andK=k=0∑∞(∣a∣∣br∣+∣s∣)k.
Fix α≥0 such that
Df(h)=o(h)andf(h)=O(hα).
Fix ϵ>0. We must find δ>0 and L>0 such that whenever ∣h∣<δ,
∣f(h)∣<Kϵ∣h∣+L∣h∣α+t.
Using that f(h)=O(hα), choose δ>0 and B>0, such that
∣f(h)∣≤B∣h∣αfor all ∣h∣<∣s∣δ.(14)
Since
af(h)+bf(rh)+f(sh)=o(h),
shrink δ, if necessary, so that for all ∣h∣<δ,
∣f(h)∣≤∣a∣1∣f(sh)∣+∣a∣∣b∣∣f(rh)∣+ϵ∣h∣.(15)
If ∣sh∣<δ (and therefore ∣rh∣<δ), then applying the same estimate to rh and sh gives
Substituting into (18) and letting L=δtB, we obtain
∣f(h)∣≤B∣h∣α(δ∣h∣)t+Kϵ∣h∣=L∣h∣α+t+Kϵ∣h∣
The case ∣a∣+1≤∣b∣ and ∣s∣+∣a∣<∣br∣
In this section we prove the following theorem.
Theorem 24. Suppose that
1<∣r∣<∣s∣ \quad and \quad ∣a∣+1≤∣b∣ \quad and \quad ∣s∣+∣a∣<∣br∣.
Then the difference
D(f,h)=af(h)+bf(rh)+f(sh)
has the MZ-property.
This time, we do not have a proof that handles both the rational and irrational cases simultaneously. We will therefore split them. First, the rational case:
Theorem 25. Suppose that
1<∣r∣<∣s∣ \quad and \quad ∣a∣+1≤∣b∣ \quad and \quad ∣s∣+∣a∣<∣br∣.
Suppose also that log∣s∣log∣r∣∈Q. Then the difference
D(f,h)=af(h)+bf(rh)+f(sh)
has the MZ-property.
Proof. Let log∣s∣log∣r∣=jk where k,j are positive integers. Let λ=∣s∣1/j, so that λj=∣s∣ and λk=∣r∣. By Corollary 12, it is enough to show that no trinomial of the form
P(z)=±zj±bzk±a
can have zeros in the annulus 1<∣z∣≤λ.
If P(z)=0 then ∣bzk∣=∣±zj±a∣≤∣z∣j+∣a∣. In other words, φ(x)=xj−∣b∣xk+∣a∣≥0 when x=∣z∣. We will show this can’t happen for x∈(1,λ]. By Descartes Rule of Signs, φ has at most two positive real roots. To estimate them,
φ(0)=∣a∣>0.
φ(1)=1−∣b∣+∣a∣≤0
φ(λ)=∣s∣−∣b∣∣r∣+∣a∣<0.
and for large enough x,
φ(x)>0.
So, φ(x) has a root in (0,1] and another root in (λ,∞) and is strictly negative in the region (1,λ].
We now consider the same theorem for the irrational case. To prove this theorem, it will be convenient to use the following lemma. Examples 13 and 14 demonstrate that this next lemma is false in the rational case, explaining why we split into cases.
Lemma 26. Let
D1(f,h)=af(h)+bf(rh)+f(sh)
and
D2(f,h)=cf(h)+df(rh)+f(sh)
be two irrational differences with the same nodes. Suppose ∣a∣=∣c∣ and ∣b∣=∣d∣. Then D1 has the MZ-property if and only if D2 has the MZ-property.
Proof. Suppose that D1 does not have the MZ-property. Then for some function f:R→C we have that D1(f,h)=o(h), f=o(1), but f=o(h). We will finish the proof by finding a function g:R→C such that D2(g,h)=o(h), g=o(1), but g=o(h).
Since ∣a∣=∣c∣, choose θ1∈R such that c=aeiθ1. Since ∣b∣=∣d∣, choose θ2∈R such that deiθ2=beiθ1.
Define
H={rnsm:n,m∈Z}.
Since log∣s∣log∣r∣∈/Q, every element of H has a unique representation of the form rnsm.
Using that f is not o(h), let hk→0 be such that f(hk)=o(hk). Consider the collection of cosets {hkH}, and note that this collection may have duplicates. If h∈⋃hkH, we will define g(h) as follows. Find the first k such that h∈hkH, and express h as h=hkrnsm. Then let g(h)=ei(mθ1+nθ2)f(h). If h∈/⋃hkH, simply let g(h)=0.
Since ∣g(h)∣≤∣f(h)∣, we get g=o(1). Since ∣g(h)∣=∣f(h)∣, for h∈⋃hkH, we get g(hk)=o(hk). Therefore, g=o(h). It remains to show that D2(g,h)=o(h).
If h∈⋃hkH, then the first k such that h∈hkH is also the first k such that rh∈hkH and the first k such that sh∈hkH. Then writing h=hkrnsm we have
This lemma will be combined with the following theorem. Although it is only needed in the irrational case, the proof of the next theorem goes through for either rational or irrational cases.
Theorem 27. Let c,d be real numbers such that
0<∣c∣<1<∣d∣,
and let c1>0,d1>0 be real numbers such that
c1+d1≤1andc1∣c∣+d1∣d∣<1.
Then
D(f,h)=f(h)−c1f(ch)−d1f(dh)
has the MZ-property.
Proof. Set
λ:=∣d∣d11−c1∣c∣andρ:=log∣d∣logλ.
By hypothesis, λ>1 and ρ>0.
Suppose D(f,h)=o(h) and f(h)=o(1). We will show f(h)=o(h). Suppose on the contrary,
hf(h)→0.
Replacing f with −f if necessary, we may assume that there exists α>0 such that for arbitrarily small h,
∣h∣f(h)>α.
Since f(h)=o(1) and D(f,h)=o(h), there exists β>0 such that whenever ∣h∣<β we have
But this contradicts the choice of β, because ∣hi∣<β and because ∣f(h)∣<1 whenever ∣h∣<β. This contradiction establishes Claim 6.
Claim 7.ϵ0<f(hi)<1 for all i.
Proof of Claim 7: By Claim 6, ∣hi∣<β for all i. Therefore, by choice of β, f(hi)<1 for all i, and by Claim 4, f(hi)>ϵ0 for all i.
Claim 8.∣hi∣→0.
Proof of Claim 8: Suppose first that Case 2 occurs infinitely many times. Then Claims 1 and 5 imply G(hi)→∞. By definition, G(hi)=∣hi∣f(hi)−ϵ0 and by Claim 7, ∣f(hi)∣<1. So, this is possible only if ∣hi∣→0.
Now suppose that Case 2 occurs only finitely many times. Then from some point onward only Case 1 occurs, so
hi+1=chi
for all sufficiently large i. Hence
∣hi+1∣=∣c∣∣hi∣
for all sufficiently large i, and since 0<∣c∣<1, it follows that hi→0. This finishes the proof of Claim 8.
To finish the proof of the theorem, note that Claims 7 and 8 directly contradict that f(h)=o(1). Therefore our assumption was false, and hence
f(h)=o(h).
We are now ready to handle the irrational case. This next theorem together with Theorem 25 establishes Theorem 24.
Theorem 28. Let D(f,h)=af(h)+bf(rh)+f(sh) be an irrational difference with 1<∣r∣<∣s∣. Suppose that
Since 0<t0≤1, f(h)=o(1) and f(h)=o(h). For h∈H, and hence both rh∈H and sh∈H,
D(f,h)=(a+b∣r∣t0eiθ2+∣s∣t0eiθ1)f(h)=0
by (1). For h∈/H, and hence both rh∈/H and sh∈/H,
D(f,h)=0.
Thus
D(f,h)≡0for all h∈R.
The following examples show that Theorem 29 can go either way in the rational case.
Example 30. Let D(f,h)=f(h)+2f(4h)+f(8h), with a=1,b=2,r=4,s=8. Then D(f,h) is rational with ∣a∣+∣b∣>1 and ∣∣a∣−∣br∣∣=7<∣s∣. The inequalities of Theorem 29 are satisfied. Since the nodes are positive, there is only one characteristic polynomial, x3+2x2+1. The roots are approximately −2.21, 0.1+0.67i, and 0.1−0.67i. The absolute values of its roots are 2.21 and 0.67, which are outside of (1,2] so by Theorem 12, D(f,h) has the MZ-property.
Example 31. Let D(f,h)=f(h)−3f(2h)+f(8h). Then D(f,h) is rational with a=1,b=−3,r=2,s=8. Then ∣a∣+∣b∣>1 and ∣∣a∣−∣br∣∣=5<∣s∣. The inequalities of Theorem 29 are satisfied. Since the nodes are positive, there is only one characteristic polynomial, x3−3x+1. One of its roots, 1.53, is in the region (1,2]. So, by Theorem 12, D(f,h) does not have the MZ-property.
Theorem 32. Suppose that
1+∣a∣>∣b∣and∣s∣<∣b∣∣r∣−∣a∣.
Then for the difference
D(f,h)=∣a∣f(h)−∣b∣f(rh)+f(sh)
there exists a complex-valued function f such that
Theorem 33.LetD(f,h)=af(h)+bf(rh)+f(sh)be an irrational difference with distinct nodes, 1,r,s where 1≤∣r∣≤∣s∣. If D has the MZ-property then it satisfies at least one of the following conditions:
∣a∣+∣b∣≤1
∣br∣+∣s∣<∣a∣
∣a∣+1≤∣b∣and∣s∣+∣a∣<∣br∣.
Proof. Suppose D(f,h)=af(h)+bf(rh)+f(sh) is irrational and does not satisfy any of the three conditions. We will show it does not have the MZ-property. Since D(f,h) is irrational, we have 1<∣r∣<∣s∣.
Case 1: ∣a∣+∣b∣>1, ∣br∣+∣s∣≥∣a∣, and ∣s∣+∣a∣≥∣br∣. Then ∣s∣≥∣br∣−∣a∣ and ∣s∣≥∣a∣−∣br∣. So, ∣s∣≥∣a∣−∣br∣. By Theorem 29, D(f,h) does not have the MZ-property.
Case 2: ∣a∣+∣b∣>1, ∣br∣+∣s∣≥∣a∣, and ∣s∣+∣a∣<∣br∣. Then ∣a∣+1>∣b∣, so by Theorem 32, the difference ∣a∣f(h)−∣b∣f(rh)+f(sh) fails the MZ-property. Then by Lemma 26, D(f,h)=af(h)+bf(rh)+f(sh) also fails the MZ-property.
References
[1]J. M. Ash, S. Catoui, and M. Csörnyei, Generalized vs. ordinary differentiation, Proc. Amer. Math. Soc. 145 (2017), 1553–1565.DOI
[2]J. M. Ash, S. Catoui, and H. Fejzić, Gaussian Riemann derivatives, Israel J. Math. 255 (2023) (1), 177–199.arxiv.org/abs/2211.09209
[3]J. M. Ash, S. Catoui, and H. Fejzić, Two pointwise characterizations of the Peano derivative, Results Math. 79 (2024) (7), Article No. 251, 17pp.arxiv.org/abs/2209.04088
[4]S. Catoui, and H. Fejzić, NonEquivalence Between Peano and Riemann derivatives, Israel J. Math. 274 (2026), no. 1, 407–419.
[5]S. Catoui and H. Fejzić, Problem for MZ differentiable functions. 1-14. (Pre-print)
[6]H. Fejzić, The Marcinkiewicz–Zygmund Property for Riemann Differences with Geometric Nodes, Real Anal. Exchange 50(2) (2025), 445–468.arxiv.org/abs/2507.11463
[7]J. Marcin kiewicz and A. Zygmund, On the differentiability of functions and summability of trigonometric series, Fund. Math. 26 (1936), 1–43.DOI