Right now the assessment script is sort of all over the place in almost like a scratch pad
Hardest part in streamlining will be deciding want dimensions (lat, lon, substrat, species, year, rep) to leave intact, and which to compress with summary statistics. Right now I have some nice space+time figures for Nsite (quick aside: I should change way I get Nsite.obs), which hopefully will be of a more reasonable size when I use 6 years instead of 12. (On the other hand, this means I won't be able to make the exact same figure for the empirical data set [which has far more years]).
Need to spend time thinking about the best approach to assess model performance. There's some of the items in #87, but that doesn't really say how each of those items should be assessed. Hmmm.
I.e., some of the formatting in #87
Right now the assessment script is sort of all over the place in almost like a scratch pad
Hardest part in streamlining will be deciding want dimensions (lat, lon, substrat, species, year, rep) to leave intact, and which to compress with summary statistics. Right now I have some nice space+time figures for Nsite (quick aside: I should change way I get
Nsite.obs), which hopefully will be of a more reasonable size when I use 6 years instead of 12. (On the other hand, this means I won't be able to make the exact same figure for the empirical data set [which has far more years]).Need to spend time thinking about the best approach to assess model performance. There's some of the items in #87, but that doesn't really say how each of those items should be assessed. Hmmm.