Need an optional "blank_dc" function

UCBerkeleySETI / hyperseti

A SETI / technosignature search code to find intelligent life beyond Earth

https://hyperseti.readthedocs.io

10 stars 4 forks source link

Need an optional "blank_dc" function #52

Closed texadactyl closed 1 year ago

texadactyl commented 2 years ago

Once the huge DC spike in the Voyager .h5 file was avoided, hyperseti behaved a lot better. See issue #51.

texadactyl commented 2 years ago

For the antenna arrays (E.g. MeerKAT), this is probably not a requirement.

telegraphic commented 2 years ago

Hi @texadactyl, working on hyperseti over my Easter break (just merged in your PR)

texadactyl commented 2 years ago

@telegraphic Saw: "DC bin needs to be blanked. This is now done with sk_flag". But, I thought that "SK" referred to spectral kurtosis. How does that fulfill the smoothing of the center of the coarse channel for those situations where it is needed?

telegraphic commented 1 year ago

between SK flagging and blank_extrema functions, DC bin is no longer an issue when calculating SNR and mean for normalization. As such special treatment of center channel is not required.