search

James-Thorson / EOFR

Package to conduct EOF regression
GNU General Public License v3.0
0 stars 1 forks source link

Spatial info for fishery data #4

Open egkennedy opened 4 years ago

egkennedy commented 4 years ago

Hi Jim. I'm clearly missing something big here. Where do you put in coordinates or spatial information for your fishery data in the code? It's clear how spatial information is plugged into the physical submodel, but there doesn't seem to be anywhere that spatial information about the the fishery data itself goes. I would have assumed make_data(), but that only seems to accept the spatial_list() outputs, which are all based on the physical submodel inputs.

James-Thorson-NOAA commented 4 years ago

I'm sorry, we might be able to talk more clearly if I know the particulars of what you're trying to do. But in the published example, the physics has spatially referenced measurements, and the biology is some spatially integrated process -- it is not associated with any single location. Said another way, EOFR relates modes of variability in spatio-temporal physics to a purely temporal biological process (i.e., integrates across unknown behavioral responses of biology in space). So the biology (time-series) has no spatial coordinates.

egkennedy commented 4 years ago

Thank you - I definitely had not picked up on that aspect of EOFR. I am working with Jessica Cross, Darren Pilcher, and Elizabeth Siddon on exploring the relationship of the new carbonate outputs of the Bering Sea NPZ model to fisheries data. EOFR and VAST are clear mechanisms to do that, but I am relatively new to fishery statistics and extremely new to the EOFR and VAST packages, so I have struggled some to understand the limits and capabilities of both techniques.

James-Thorson commented 4 years ago

Sorry I remember this now ;) you wanna email me to set up a chance to chat by video

On Thu, Aug 20, 2020, 4:02 PM egkennedy notifications@github.com wrote:

Thank you - I definitely had not picked up on that aspect of EOFR. I am working with Jessica Cross, Darren Pilcher, and Elizabeth Siddon on exploring the relationship of the new carbonate outputs of the Bering Sea NPZ model to fisheries data. EOFR and VAST are clear mechanisms to do that, but I am relatively new to fishery statistics and extremely new to the EOFR and VAST packages, so I have struggled some to understand the limits and capabilities of both techniques.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/James-Thorson/EOFR/issues/4#issuecomment-677948038, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB46UTMGUMFORKQKWOZDBFDSBWTO7ANCNFSM4QGVJX4Q .