Skip to content

Local covariance matrix #156

Description

@Aseman7

Hello,

Please have a look at Figure 5 of Czekala et al. 2015:

Screenshot 2023-04-13 at 11 03 53

It explains how the addition of local kernel can be so useful in some cases. My question is: how can I add an outlier line (for example, 5202.2 Angestrum) into code?

As the Starfish documentation says, I have to define its kernel by its hyperparameters:

log_amp: The natural logarithm of the amplitude of the kernel
mu: The location of the local kernel
log_sigma: The natural logarithm of the standard deviation of the kernel

Does it mean that I use the following definition in the SpectrumModel function as:

local_cov = dict(log_amp:?, mu: 5202.2, log_sigma:?)

If yes, please guide me on how to select "log_amp" and "log_sigma". Does "mu" value is correct?

And how to define them in the prior values:

"local_cov:log_amp": st.norm(?),
"local_cov:log_sigma": st.norm(?),
"local_cov:mu": st.norm (5222.1, 0.1)

Thanks a lot for any help in advance.

Activity

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

Metadata

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