Add thicknessOf()

[brownag]

Adds a commonly sought-after operation thicknessOf() allowing for calculation of cumulative or "minmax" thickness of horizons matching some criteria

This function calculates the cumulative (method="cumulative", default) or maximum difference between (method="minmax") horizons within a profile that match a defined pattern (pattern) or, more generally, any set of horizon-level logical expressions encoded in a function (FUN).

TODO:

• add tests
• add column prefix (optional) for easy joining of various thickness summaries sequentially to site table
• return pattern (like depthOf()) I actually dont think this is necessary
• allow for customization of min/max variable names (default to horizonDepths(x))

EDIT: fix for "minmax"

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

# cumulative thickness of horizon designations matching "Bt"
thicknessOf(jacobs2000, "Bt")
#>     id thickness
#> 1 92-1       110
#> 2 92-2        99
#> 3 92-3       111
#> 4 92-4         0
#> 5 92-5        26
#> 6 92-6        64
#> 7 92-7         0

# maximum bottom depth minus minimum top depth of horizon designations matching "Bt"
thicknessOf(jacobs2000, "Bt", method = "minmax")
#>     id tmin tmax thickness
#> 1 92-1   43  153       110
#> 2 92-2   46  145        99
#> 3 92-3   64  175       111
#> 4 92-4  Inf -Inf         0
#> 5 92-5  109  135        26
#> 6 92-6  104  168        64
#> 7 92-7  Inf -Inf         0

# cumulative thickness of horizon designations matching "A|B"
thicknessOf(jacobs2000, "A|B")
#>     id thickness
#> 1 92-1       131
#> 2 92-2       117
#> 3 92-3       136
#> 4 92-4        20
#> 5 92-5        54
#> 6 92-6       110
#> 7 92-7        43

# maximum bottom depth minus minimum top depth of horizon designations matching "A|B"
thicknessOf(jacobs2000, "A|B", method = "minmax")
#>     id tmin tmax thickness
#> 1 92-1    0  156       156
#> 2 92-2    0  145       145
#> 3 92-3    0  175       175
#> 4 92-4    0   20        20
#> 5 92-5    0  135       135
#> 6 92-6    0  168       168
#> 7 92-7    0  140       140

# note that the latter includes the thickness of E horizons between the A and the B

# when using a custom function (be sure to accept ... and consider the effect of NA values)

# calculate cumulative thickness of horizons containing >18% clay
thicknessOf(jacobs2000, FUN = function(x, ...) !is.na(x[["clay"]]) & x[["clay"]] > 18)
#>     id thickness
#> 1 92-1       170
#> 2 92-2       167
#> 3 92-3        81
#> 4 92-4         0
#> 5 92-5         0
#> 6 92-6        49
#> 7 92-7         0

[dylanbeaudette]

Nice idea. I'll do some testing this week,

[dylanbeaudette]

I can't remember if we wanted to default to guessing (across all functions) when horizon names aren't specified or set in metadata. Currently there is a mixture of approaches across aqp.

[brownag]

I know you have more recently decided to not guess (which has broken or changed output in several places), but I don't think there is a rule we agreed on. If you are making one then we can start a change here.

The functions for guessing horizon designation, texture class names, and generic attributes are used in several places (pretty much all "my" functions I suppose) I could replace all uses of guessxx() with the metadata getter, and required=TRUE argument, but for now this is consistent with many of my other function usages. Let's address that (deprecating the 3 guess functions) in a separate issue or PR.

[brownag]

I have started to address this in #309, and will make a corresponding fix in this PR. Thanks for prodding me on this, there was lots of room to improve consistency and I am happy to tackle it across all functions.

[dylanbeaudette]

Tiny request: can you please add in a couple of comments so that I can keep track of the logic. Thanks.

[brownag]

Tiny request: can you please add in a couple of comments so that I can keep track of the logic. Thanks.

Done in a466959

[brownag]

Merging this now, can make any adjustments/additional method demonstrations in future vignette or examples