The maximum width of an arbitrary rooted tree is a balance index and is defined as the number of vertices at its most abundant level. It is a well-studied parameter in computer science (e.g., Drmota, Hwang, 2005; Devroye, Hwang, 2006), where it is often simply referred to as the width of a tree. More recently, it was suggested as a tree shape statistic for phylogenetic trees by Colijn, Gardy, 2014. It can be calculated using the function maxWidth from our R package treebalance.
Definition (Colijn, Gardy, 2014): The maximum width of a tree
with height
is defined as
where is the number of vertices
that have depth
.
Computation time: For every tree , the maximum width
can be computed in time
(Walter, 2022, Proposition 1; see also Fischer et al., 2023, Proposition 23.16).
Recursiveness: Open problem.
Locality: The maximum width is not local (Walter, 2022, Proposition 3; see also Fischer et al., 2023, Proposition 23.17).
Maximal value and trees with maximal value on : An upper bound for the maximum width for trees
which is tight for all
as well as a characterization of the corresponding maximal trees have already been established (see Fischer et al., 2023, Theorem 23.10).
Number of trees with maximal value on : Open problem.
Maximal value on : Partially open problem. An upper bound for the maximum width for binary trees
which is tight for all
has already been established (see Fischer et al., 2023, Theorem 23.11). An explicit formula is — to our knowledge — not yet known.
(Number of) trees with maximal value on : Partially open problem. The number of binary trees
with maximal maximum width for all
with
has already been established (see Fischer et al., 2023, Corollary 23.1). If
is not a power of two, the maximal tree(s) have — to our knowledge — not been analyzed yet.
Minimal value and (number of) trees with minimal value on and
: A lower bound for the maximum width for trees
which is tight for all
as well as the (number of) corresponding minimal trees have already been established (see Fischer et al., 2023, Theorem 23.12).
Expected value and variance under the Yule model: Open problem.
Expected value and variance under the uniform model: Open problem.