auto_rate - duplicate results in summary

[nicholascarey]

auto_rate often has multiple exact same results in summary table. For example (not the only example in this dataset):

res <- auto_rate(sardine.rd) 
res$summary[17:21]

   row endrow time endtime intercept_b0      rate_b1   rsq
1: 582   1988  581    1987     95.72169 -0.001037267 0.961
2: 570   1992  569    1991     95.71753 -0.001034097 0.962
3: 582   1988  581    1987     95.72169 -0.001037267 0.961
4: 582   1988  581    1987     95.72169 -0.001037267 0.961
5: 582   1988  581    1987     95.72169 -0.001037267 0.961

I guess these are distinct results in the KD analysis, but not much point having them in the summary when they are literally identical. unique() on the df removes them.

They get removed as first step in new fn subset_rate, but maybe do that in auto_rate first?

[nicholascarey]

Partially addressed in 0005733

Added density to the summary table for linear method. Users can see that they have a different density score, but they still seem redundant, so would be a good idea to just remove them?

Alternatively, does the fact that some exact same regions return multiple, different results in the KDE analysis mean tehy should be ranked higher? That is, combined in some way?

[nicholascarey]

Fixed in 5f19afb

Duplicates now removed in auto_rate linear method.

Would still like to know why they happen at all, or if they should be higher ranked because they are identified multiple times. @januarharianto ?