Skip to content

Persistence weight function and negative values in persistence image #57

Open
@davafe

Description

The 'persistence' weight function in PersistenceImager is the one defined here https://jmlr.org/papers/volume18/16-337/16-337.pdf (page 9) that is linear on persistence coordinate and goes from 0 if persistence coordinate is 0 to 1 if persistence coordinate is the maximum?

I have obtained negative values in a persistent image (they were very close to 0). How is that possible? I have used Gaussian kernel and 'persistence' weight function, so the values in a persistence image are sums of the product of nonnegative functions.

Activity

changed the title Persistence weight function and Persistence weight function and negative values in persistence image on Jun 1, 2022
catanzaromj

catanzaromj commented on Oct 23, 2023

@catanzaromj
Contributor

@davafe I'm not sure how that is possible. Could you provide a minimal reproducible example of the code that produced the negative values so we can take a look and debug?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      Persistence weight function and negative values in persistence image · Issue #57 · scikit-tda/persim