ideas related to #43

misc/print-ST.R

@@ -9,36 +9,45 @@ library(data.tree)
 ## TODO:
 # * add subgroup acreage
 # * 13th ed.
-# * compact style of dir_tree(), but in proper order
 # * add series count to subgroups
 
+# 12th ed.
 data("ST")
 
+# latest SC DB
 sc <- get_soilseries_from_NASIS()
 
+# susbet / rename columns for simpler joining
 sc <- sc[, c('soilseriesname', 'taxclname', 'taxsubgrp')]
 names(sc) <- c('series', 'family', 'subgroup')
 
+# combine ST hierarchy + series by subgroup
+# this will introduct family details
 z <- merge(ST, sc, by.x = 'subgroup', all.x = TRUE, sort = FALSE)
 head(z)
 
+# normalization via lower case
 z$family <- tolower(z$family)
 z$series <- tolower(z$series)
 
+# remove subgroup component of family spec
 z$f <- NA_character_
 for(i in 1:nrow(z)) {
   z$f[i] <- gsub(pattern = z$subgroup[i], replacement = '', z$family[i], fixed = TRUE)
 }
 
+# remove white space
 z$f <- trimws(z$f, which = 'both')
 
+# ok
 head(z)
 
 
 # ordering used by 'Keys
 z <- z[order(z$code, method = 'radix'), ]
 
-
+## hack #1: use directory listing for a compact representation
+# note: order isn't correct
 td <- tempdir()
 unlink(file.path(td, 'ST'), recursive = TRUE)
 
@@ -70,16 +79,20 @@ for(i in 1:nrow(z)) {
   dir.create(path = fp, recursive = TRUE)
 }
 
+
+# dump output file text file
 setwd(td)
 
 sink('e:/temp/st12.txt')
 dir_tree(file.path('ST'))
 sink()
 
+# cleanup
 unlink(file.path(td, 'ST'), recursive = TRUE)
 
 
-
+## hack #2: use directory listing for a compact representation
+##          prefix with taxon codes for correct ordering
 for(i in 1:nrow(z)) {
 
   # account for subgroups without series
@@ -110,25 +123,35 @@ for(i in 1:nrow(z)) {
   dir.create(path = fp, recursive = TRUE)
 }
 
+# save output to text file
 setwd(td)
 
 sink('e:/temp/st12-codes.txt')
 dir_tree(file.path('ST'))
 sink()
 
+# cleanup
 unlink(file.path(td, 'ST'), recursive = TRUE)
 
 
 
+## correct ordering via data.tree, as long as order of `z` is correct
+# less compact, but doesn't require crazy file system manipulation
 
+# required columns only, smaller data.tree
 v <- c('order', 'suborder', 'greatgroup', 'subgroup', 'f', 'series', 'path')
+
+# init data.tree object
 z$path <- sprintf("ST/%s/%s/%s/%s/%s/%s", z$order, z$suborder, z$greatgroup, z$subgroup, z$f, z$series)
 n <- as.Node(z[, v], pathName = 'path')
 
+## missing family / series result in an ugly tree, prune accordingly
 
 # prune missing family / series
 pf <- function(i) {
 
+  # NA due to left join
+  # note odd approach required, matching to 'NA' vs. is.na()
   if(GetAttribute(i, 'name') == 'NA') {
     return(FALSE)
   } else {
@@ -137,10 +160,13 @@ pf <- function(i) {
 
 }
 
+# dump to text file
 options('max.print' = 1e7)
 sink('e:/temp/st12-DT.txt')
 print(n, limit = NULL, pruneFun = pf)
 sink()
 
 options('max.print' = 1000)
 
+
+