search

anderdyl / objectiveMapping

functions to combine different FRF surveys into an integrated topo map
1 stars 0 forks source link

acceptable threshold for "good" vs "bad" data #3

Open SBFRF opened 4 years ago

SBFRF commented 4 years ago

https://github.com/anderdyl/objectiveMapping/blob/d3b4255b9d8456c663a52432fd44dcf7addb7ddc/operation_FRF_duneLidarDEM_bathy_interp.py#L157

Hey @anderdyl if i'm understanding correctly, this is the portion of code that defines a threshold for what would be considered too much error (NMSEET) to incorporate into final grid.

  1. nmseEst - estimated normalized mean square error from the map_interp function ??
  2. how did you set this 0.2 threshold? was it recommended or educated guess/experience ?
anderdyl commented 3 years ago

Pretty sure the 0.2 came from Bonny... but I do think that it has some dependence on the resolution of the data and the resolution of the grid (both affecting the number of observations in each bin). I think a sensitivity test to that parameter would make sense.