From 16faf22b9feab26f1d27e6152298e06a4b284b7d Mon Sep 17 00:00:00 2001
From: Taha Ahmed <taha@mkem150>
Date: Tue, 27 Dec 2011 11:21:31 +0100
Subject: [PATCH] numbers2words() converts numbers into their spelled names.
 capitalize is...

... a simple function for capitalizing the first letter of a string.
It should work well in conjunction with numbers2words() at the beginning
of sentences.
---
 capitalize.R    | 17 +++++++++++
 numbers2words.R | 75 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 92 insertions(+)
 create mode 100644 capitalize.R
 create mode 100644 numbers2words.R

diff --git a/capitalize.R b/capitalize.R
new file mode 100644
index 0000000..431b3d8
--- /dev/null
+++ b/capitalize.R
@@ -0,0 +1,17 @@
+##################################################
+################## capitalize ####################
+##################################################
+capitalize <- function(x) {
+   ## Description:
+   ##   Capitalizes the first letter of a string
+   ##   == This function was inspired by the function supplied in the base R doc for chartr()
+   ## Usage:
+   ##   capitalize(string)
+   ## Arguments:
+   ##   x: a string or vector of strings
+   ## Value:
+   ##   A string or vector of strings
+   #
+   paste(toupper(substring(x, 1, 1)), substring(x, 2), 
+         sep = "")
+}
diff --git a/numbers2words.R b/numbers2words.R
new file mode 100644
index 0000000..8c3b438
--- /dev/null
+++ b/numbers2words.R
@@ -0,0 +1,75 @@
+##################################################
+################ numbers2words ###################
+##################################################
+numbers2words <- function(x, billion = c("US", "UK"), and = if (billion == "US") "" else "and") {
+   ## Description:
+   ##   Converts a number into its corresponding words in English
+   ##   == THIS FUNCTION WAS PUBLISHED IN: R-News, vol 5, iss 1, May 2005, pp. 51
+   ##   == Original author: John Fox
+   ##   == Department of Sociology
+   ##   == McMaster University
+   ##   == Hamilton, Ontario
+   ##   == Canada L8S 4M4
+   ##   == 905-525-9140x23604
+   ##   == http://socserv.mcmaster.ca/jfox 
+   ##   == http://cran.csiro.au/doc/Rnews/Rnews_2005-1.pdf
+   ##   == http://finzi.psych.upenn.edu/R/Rhelp02a/archive/46843.html
+   ## Usage:
+   ##   numbers2words(number, billion-word, and-or-not)
+   ## Arguments:
+   ##   x: number
+   ##   billion: follow either US or UK usage rules
+   ##   and: follows the choice set in billion arg
+   ## Value:
+   ##   Dataframe with the following columns:
+   ##   $ sampleid        : chr
+   #
+   billion <- match.arg(billion)
+   trim <- function(text) {
+      gsub("(^\ *)|((\ *|-|,\ zero|-zero)$)", "", text)
+   }
+   makeNumber <- function(x) as.numeric(paste(x, collapse = ""))
+   makeDigits <- function(x) strsplit(as.character(x), "")[[1]]
+   helper <- function(x) {
+      negative <- x < 0
+      x <- abs(x)
+      digits <- makeDigits(x)
+      nDigits <- length(digits)
+      result <- if (nDigits == 1) as.vector(ones[digits])
+      else if (nDigits == 2)
+      if (x <= 19) as.vector(teens[digits[2]])
+      else trim(paste(tens[digits[1]], "-", ones[digits[2]], sep=""))
+      else if (nDigits == 3) {
+         tail <- makeNumber(digits[2:3])
+         if (tail == 0) paste(ones[digits[1]], "hundred")
+         else trim(paste(ones[digits[1]], trim(paste("hundred", and)),
+         helper(tail)))
+      } else {
+         nSuffix <- ((nDigits + 2) %/% 3) - 1
+         if (nSuffix > length(suffixes) || nDigits > 15)
+         stop(paste(x, "is too large!"))
+         pick <- 1:(nDigits - 3 * nSuffix)
+         trim(paste(helper(makeNumber(digits[pick])), suffixes[nSuffix], helper(makeNumber(digits[-pick]))))
+      }
+      if (billion == "UK") {
+         words <- strsplit(result, " ")[[1]]
+         if (length(grep("million,", words)) > 1)
+         result <- sub(" million, ", ", ", result)
+      }
+      if (negative) paste("minus", result) else result
+   }
+   opts <- options(scipen = 100)
+   on.exit(options(opts))
+   ones <- c("zero", "one", "two", "three", "four", "five", "six", "seven", "eight", "nine")
+   teens <- c("ten", "eleven", "twelve", "thirteen", "fourteen", "fifteen", "sixteen", " seventeen", "eighteen", "nineteen")
+   names(ones) <- names(teens) <- 0:9
+   tens <- c("twenty", "thirty", "forty", "fifty", "sixty", "seventy", "eighty", "ninety")
+   names(tens) <- 2:9
+   suffixes <- if (billion == "US") {
+         c("thousand,", "million,", "billion,", "trillion,")
+      } else {
+         c("thousand,", "million,", "thousand million,", "billion,")
+      }
+   x <- round(x)
+   if (length(x) > 1) sapply(x, helper) else helper(x)
+}