Revisit how to trim years/strata for Gulf of Mexico

[mpinsky]

We currently trim to strata sampled at least 10 times between 2008 and 2018, and we only use data from >2008. This ends up losing a lot of data (about 25 years).

Can we instead trim to strata sampled consistently through time?

[df511]

If we trim from 2008 onward and to strata with more than 10 years coverage, we lose ~6% of the data (plot 1, top). If we only trim by strata surveyed for more than 15 years, leaving pre-2008 data in, we lose ~18% of data and coverage is still fairly sparse (plot 2, bottom).

[df511]

@mpinsky I was just re-running the full script to make sure there are no lingering bugs, and found that the code used to trim the Gulf of Mexico data and generate the above plot no longer produces the same plot. I must not have been overwriting the lower plots, because all of the percentages removed are consistent with the old plot (20.3% of data removed), but this makes more sense in looking at the original data. What do you think about removing the whole eastern half of the Gulf of Mexico data? Is it still worth it? I would likely still have to trim more to get more consistent temporal coverage. Showing this plot for consistency, but I'll drop a plot with better temporal coverage below.

[df511]

Here's a better filter. We lose 23% of the original data, but get decent temporal coverage. @mpinsky

[mpinsky]

Good catch, and I now have a memory of seeing that before. I’m ok losing the spatial coverage in this case I’m favor of the longer time series.

On Tue, Apr 6, 2021 at 5:07 PM df511 ***@***.***> wrote: @mpinsky <https://github.com/mpinsky> I was just re-running the full script to make sure there are no lingering bugs, and found that the code used to trim the Gulf of Mexico data and generate the above plot no longer produces the same plot. I must not have been overwriting the lower plots, because all of the percentages removed are consistent with the old plot (20.3% of data removed), but this makes more sense in looking at the original data. What do you think about removing the whole eastern half of the Gulf of Mexico data? Is it still worth it? I would likely still have to trim more to get more consistent temporal coverage. Showing this plot for consistency, but I'll drop a plot with better temporal coverage below. [image: image] <https://user-images.githubusercontent.com/30601505/113778198-1dbf5180-96fa-11eb-8f91-140e8b473b21.png> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#124 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABQTJZDURHQN2A3D7EPITP3THNZYDANCNFSM4MFI24IA> .

-- Please excuse, sent from a device with tiny keys...