DistanceDevelopment / mrds

R package for mark-recapture-distance-sampling analysis
GNU General Public License v3.0
4 stars 4 forks source link

Estimates by sample (Nhat.by.sample) error when point transects visited multiple times #100

Open erex opened 8 months ago

erex commented 8 months ago

Thanks to a comment to the list density per point transect station appears not to incorporate information about multiple visits (contained in Effort). Consequently, density estimates are likely to be incorrect. There may be broader implications.

Issues with Nhat.by.sample were previously raised in November 2022

See this bit of output, with different transects visited different number of times: image

At some time in the past, mrds code did consider effort when computing covered area as shown in this similar bit of output image

Per-transect density estimates differ by >an order of magnitude in the two results.

erex commented 8 months ago

Further details from the original poster:

Thanks for the response. I believe that this is definitely at the heart of the issue because when I went through and calculated what estimates the code should be populating if it had the sample area calculated correctly, the site/station level are coming out to what we are expecting. Also the region level is being estimated correctly (because it is based on the full area) so I know that the way the code is not calculating the sample.area is where the issue lies (plus we have been running this for several years and besides us regrouping our site/station we haven't changed anything else except adding years of data). Will you please let me know once that code is fixed for the ds function?