mutate_profile: dynamic column and expression upgrades

[brownag]

mutate_profile() is a safe and efficient method for doing profile level calculations using expressions involving SPC columns to produce or modify site or horizon level columns. However, it can be difficult to "program" with because of the evaluation of ... where each argument becomes a new column. Sometimes the column names need to be calculated from data, and sometimes we want to reuse an expression on several columns.

This PR is a proof of concept that adds mutate_profile_raw() which can be used if building sets of dynamic mutate expressions. Also mutate_profile() gains col_names argument for dynamic naming of columns.

Also adds examples to the mutate_profile() docs

Thanks to @natearoe for discussion on this topic inspiring these upgrades.

I still need to think about the ergonomics.

Another possible approach would be to bring {rlang} dependency back in and use tidy data masking or tidy selection, allowing for use of verbs like across() or starts_with() to identify input variables. Even if that were to be implemented, there would still likely be value in having an approach like mutate_profile_raw()

Example:

library(aqp)
#> This is aqp 2.0.3
data(jacobs2000)

set.seed(123)

# example with a dynamically named column name
x <- mutate_profile(jacobs2000, bottom - top, 
                    col_names = paste0("thk", floor(runif(1, 0, 100))))
x$thk28
#>  [1] 18 25 36 51 23  3 57 18 28 38 38 23 68 15 10 39 20 28 53 10 20 33 26 51 35
#> [26] 20 18 28 33 48 26 48 18 28 30 28 15 49 15 26  7 13 30 41  8 12

# example with dynamic number of columns and names
master_desgn <- c("O", "A", "E", "B", "C", "R", "L", "M")
thk_names <- paste0("thk_", master_desgn)

x$thk <- x$bottom - x$top

## construct an arbitrary number of expressions using variable inputs
ops <- lapply(master_desgn, function(x) substitute(sum(thk[grepl(VAR, name)], na.rm = TRUE), list(VAR = x)))
names(ops) <- thk_names

# do mutation
y <- mutate_profile_raw(x, ops)

site(y)[c(idname(y), thk_names)]
#>     id thk_O thk_A thk_E thk_B thk_C thk_R thk_L thk_M
#> 1 92-1     0    18    25   113    60     0     0     0
#> 2 92-2     0    18    28    99    68     0     0     0
#> 3 92-3     0    25    49   111     0     0     0     0
#> 4 92-4     0    20     0     0   183     0     0     0
#> 5 92-5     0    28    81    26    48     0     0     0
#> 6 92-6     0    46    86    64     0     0     0     0
#> 7 92-7     0    15    97    28    12     0     0     0