simplfyFragmentData() needs a full review about soildb

[dylanbeaudette]

The function also needs to be renamed to simplifyFragmentData--note the missing "i" in the current function name.

[dylanbeaudette]

Note that issues related to fragment content >100% aren't always data errors, they can be related to multiple pedons / KSSL data source (c/o Brian Gardner):

HI Dylan,

As you recall I was coming up with erroneously high frag contents in some horizons on Kettenbach series. After a closer look, it seems like the function is being confused because for this series there are two pedons per site. As is practice we have one pedon owned by KSSL that remains unchanging and a second owned by the SSO that is updateable. The updateable one is checked as 'representative pedon'.

This creates a situation where there are two horizons in NASIS with the same horizon labsample number. A quick look at horizons where frags were less than 100% showed several that were still double what was recorded for the horizon in NASIS (for each individual pedon/horizon)

Maybe this is the source of the problem?

[dylanbeaudette]

This function should also allow for custom specification of fragment classes as an argument.

[dylanbeaudette]

Function re-named with backwards compatibility in e2b237d.

[dylanbeaudette]

Another small issue, simplifyFragmentData fails when passed a single row with null fragvol:

d <- structure(list(phiid = 1386592L, fragvol = NA_real_, fragsize_l = NA_integer_, 
    fragsize_r = NA_integer_, fragsize_h = NA_integer_, fragshp = structure(NA_integer_, .Label = c("flat", 
    "nonflat"), class = "factor"), fraghard = structure(NA_integer_, .Label = c("noncemented", 
    "indurated", "moderately cemented", "strongly cemented", 
    "weakly cemented", "extremely weakly", "very weakly", "very strongly", 
    "weakly", "moderately", "strongly", "extremely strong", "H", 
    "S"), class = "factor")), .Names = c("phiid", "fragvol", 
"fragsize_l", "fragsize_r", "fragsize_h", "fragshp", "fraghard"
), row.names = 1049L, class = "data.frame")

simplifyFragmentData(d, 'phiid')

Error: value.var (volume) not found in input
In addition: Warning message:
 Error: value.var (volume) not found in input 

This is a data population issue, consider pre-filtering rows without data.

[dylanbeaudette]

Filtering solves the problem but we are now left with the issue of NA vs. 0 when there are no records in the phfrags table.