coordinatePrecision - deciding how precise the units are.

[SimonGoring]

Understanding coordinate precision is hard. I'd like to be able to include it, but it's challenging.

The first potential solution I used was to look at the number of decimal places. You can't do this in R, but I wrote a short function that converts the numeric to a character and then split the string at the decimal point. Doing this provides a table that shows the maximum decimal places in the site coordinates:

Decimal places	Count of Sites
0	55
1	225
2	1846
3	63
4	133
5	576
6	3057
7	32
8	25
9	26
10	10
12	255
13	2207
16	1

This should indicate that it's unreasonable to think that this is the best solution. Given that many of these coordinates are decimal interpretations of the d^o^m's" system it might make sense to look for values that divide evenly into dms systems.

I added the function dms_conv:

dms_conv <- function(x) {
  
  if (x < 0) { w <- TRUE } else { w <- FALSE }
  
  x <- abs(x)
  
  d <- x - (x %% 1)
  m <- ((x - d) * 60) - (((x - d) * 60) %% 1)
  s <- x - (d + m/60)
  
  c(ifelse(w, d, -d), m, ifelse(s < 0.5, 0, s))
}

So, now we can test whether the seconds are whole numbers or not, to get some estimate of precision. If s == 0, then we can look at the precision of the m. The coordinatePrecision field allows for repeating decimals, as recognizing precision in minutes &cetera.