The dual space of \(L^p\) of a vector measure (Q618826)

From MaRDI portal
scientific article
Language Label Description Also known as
English
The dual space of \(L^p\) of a vector measure
scientific article

    Statements

    The dual space of \(L^p\) of a vector measure (English)
    0 references
    0 references
    17 January 2011
    0 references
    Let \(\nu:\Sigma \to E\) a Banach space valued (countably additive) vector measure and consider the space \(L^1(\nu)\) of all integrable functions with respect to \(\nu\) and the space \(L_w^1(\nu)\) of weakly integrable functions with respect to \(\nu\). For \(1\leq p<\infty\), consider the spaces \(L^p(\nu)\) and \(L_w^p(\nu)\) of measurable functions \(f\) such that \(|f|^p\) is in \(L^1(\nu)\) or \(L_w^1(\nu)\), respectively. All these spaces are Banach function spaces over a fixed Rybakov control measure \(\mu\) for \(\nu\). Let \(1 < p, q < \infty\) be conjugate indexes. In this paper, the author characterizes the Köthe dual \(L^p(\nu)'\) of \(L^p(\nu)\) as a Banach function space \(E_q(\mu)\) of measurable functions that satisfy \[ \inf \sum_{k\geq1}\left\|\frac{g_k}{|R(x_k^*)|^{\frac{1}{p}}}\right\|_q<\infty. \] Here the infimum is computed over the set of all sequences \((g_k)\) of positive measurable functions satisfying \(|g|=\sum_{k\geq1}g_k\) \(\mu\)-a.e.\ pointwise and over all sequences \((x_k^*)\) in the unit ball of \(E^*\). The symbol \(R(x^*)\) denotes the Radon-Nikodým derivative of \(\langle\nu,x^*\rangle\) with respect to \(\mu\) for every \(x^*\in E^*\) and \(\| h\|_q=\big(\int |h|^q\,d\mu\big)^{\frac{1}{q}}\) for every measurable function \(h\). This description is useful and clear, and provides an interesting new tool for the study of the spaces \(L^p\) of a vector measure; the reader must take into account that these spaces are in general far from classical \(L^p\)-spaces, and are used to represent the class of all order continuous \(p\)-convex Banach function spaces with a weak unit. The author also proves that the Köthe dual space \(E_q(\mu)'\) is \(L_w^p(\nu)\), which implies that \(L^p(\nu)''=L_w^p(\nu)\). This result was already known, but the way of proving it is different. For the case \(p=1\), the author also gives some information; in this case, \(E_\infty\) is not necessarily equal to \((L^1(\nu))'\).
    0 references
    vector measure
    0 references
    function norm
    0 references
    associate space
    0 references
    order continuity
    0 references
    Fatou property
    0 references

    Identifiers