Skip to content

Unnecessary parameter constraint in the GUD #16

Description

@jaburgoyne

The GUD lists a parameter constraint that $a_0 = 0 \implies β = 1$. This constraint seems unnecessarily restrictive.

Deriving the GUD as a Weibull-transformed standard Pearson distribution as in Issue #15, for example, shows that not only $β = 1$ but also $b_0 = 0$ is sufficient to yield $a_0^{(\text{GUD})} = 0$. If one is willing to consider the GUD as a Weibull-transformed standard extended Pearson distribution, then the constraint seems wholly unnecessary: any combination of extended-Pearson and Weibull parameters such that $a_0^{(\text{ExtPearson})} - \frac{β-1}{β} b_0^{(\text{ExtPearson})} = 0$ yields $a_0^{(\text{GUD})} = 0$.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions