1. Introduction
This note is devoted to formulas for calculation of integrals over the n-dimensional hypercube centered at
and its boundary, based on integration over hyperplanar subsets of and exact for harmonic or polyharmonic functions. They are presented in Section 2 and can be considered as natural analogues on of Gauss surface and volume mean-value formulas for harmonic functions ([1] ) and Pizzetti formula [2] , ( [3] , Part IV, Ch. 3, pp. 287-288) for polyharmonic functions on the ball in Rn. Section 3 deals with the best one-sided L1-approximation by harmonic functions.
Let us remind that a real-valued function f is said to be harmonic ( polyharmonic of degree) in a given domain if and on, where is the Laplace operator and is its m-th iterate
For any set, denote by the linear space of all functions that are har- monic (polyharmonic of degree m) in a domain containing D. The notation will stand for the Lebesgue measure in.
2. Mean-Value Theorems
Let and be the ball and the hypersphere in
with center and radius r. The following famous formulas are basic tools in harmonic function theory and state that for any function h which is harmonic on both the average over and the average over are equal to.
The surface mean-value theorem. If, then
(1)
where is the -dimensional surface measure on the hypersphere.
The volume mean-value theorem. If, then
(2)
The balls are known to be the only sets in satisfying the surface or the volume mean-value theorem. This means that if is a nonvoid domain with a finite Lebesgue measure and if there exists a point
such that for every function h which is harmonic and integrable on, then is an
open ball centered at (see [4] ). The mean-value properties can also be reformulated in terms of quadrature domains [5] . Recall that is said to be a quadrature domain for, if is a connected open set
and there is a Borel measure with a compact support such that for every -
integrable harmonic function f on. Using the concept of quadrature domains, the volume mean-value property is equivalent to the statement that any open ball in is a quadrature domain and the measure is the Dirac measure supported at its center. On the other hand, no domains having “corners” are quadrature domains [6] . From this point of view, the open hypercube is not a quadrature domain. Nevertheless, it is proved in Theorem 1 below that the closed hypercube is a quadrature set in an extended sense, that is, we find explicitly a measure with a compact support having the above property with replaced by but the condition is violated exactly at the “corners” (for the existence of quadrature sets see [7] ). This property of is of crucial importance for the best one-sided L1-approximation with respect to (Section 3).
Let us denote by the -dimensional hyperplanar segments of defined by
(see Figure 1). Denote also
and. It can be calculated that
and
The following holds true.
Theorem 1 If, then h satisfies:
(i) Surface mean-value formula for the hypercube
(3)
(ii) Volume mean-value formula for the hypercube
(4)
In particular, both surface and volume mean values of h are attained on.
Proof. Set
and
Using the harmonicity of h, we get for
Hence, we have
(5)
if and
(6)
if.
Clearly, (5) is equivalent to (3) and from (6) it follows
(7)
which is equivalent to (4). □
Let. Analogously to the proof of Theorem 1 (ii), Equation (7) is generalized to:
Corollary 1 If and is such that and, then
(8)
The volume mean-value formula (2) was extended by P. Pizzetti to the following [2] [3] [8] .
The Pizzetti formula. If, then
Here, we present a similar formula for polyharmonic functions on the hypercube based on integration over the set.
Theorem 2 If, , and is such that, , then the following identity holds true for any:
(9)
where.
Proof. Equation (9) is a direct consequence from (8):
3. A Relation to Best One-Sided L1-Approximation by Harmonic Functions
Theorem 1 suggests that for a certain positive cone in the set is a characteristic set for the best one-sided L1-approximation with respect to as it is explained and illustrated by the examples presented below.
For a given, let us introduce the following subset of:
A harmonic function is said to be a best one-sided L1-approximant from below to f with respect to if
where
Theorem 1 (ii) readily implies the following ([6] [9] ).
Theorem 3 Let and. Assume further that the set belongs to the zero set of the function. Then is a best one-sided L1-approximant from below to f with respect to.
Corollary 2 If, any solution h of the problem
(10)
is a best one-sided L1-approximant from below to f with respect to.
Corollary 3 If, where and on, then is
a best one-sided L1-approximant from below to f with respect to.
Example 1 Let, and. By Corollary 2, the solution
of the interpolation problem (10) with is a best one-sided L1-
appro-ximant from below to f1 with respect to and. Since the function belongs
to the positive cone of the partial differential operator (that is,), one can compare
the best harmonic one-sided L1-approximation to f1 with the corresponding approximation from the linear sub- space of:
The possibility for explicit constructions of best one-sided L1-approximants from, is studied in [10] . The functions and, where and are the unique best one-sided L1-approximants to f1 with respect to from below and above, respectively, play the role of basic error functions of the cano- nical one-sided L1-approximation by elements of. For instance, can be constructed as the unique interpolant to f1 on the boundary of the inscribed square and
(Figure 2).
Example 2 Let, and. The solution
of (10) with is a best one-sided L1-approximant from
below to with respect to and. It can also be verified that (see Figure 3).
Remark 1 Let is such that, , and, on. It follows from (8) that Theorem 3 also holds for the best weighted L1-approximation from below with respect to with weight. The smoothness requirements were used for brevity and wherever possible they can be weakened in a natural way.