Quantargo exercise - modify existing package

NAMESPACE

@@ -1 +1,2 @@
 exportPattern("^[[:alpha:]]+")
+importFrom("stats", "quantile")

R/meanimpute.R

@@ -1,4 +1,5 @@
 #' Meanimputation
+#' @param x A vector
 #' @export
 meanimpute <- function(x) {
   x[is.na(x)] <- mean(x, na.rm = TRUE)

R/transform_log.R

@@ -0,0 +1,19 @@
+#' transform_log 
+#' Transform numerical values into their log values
+#' @param x A vector
+#' @return logarithm of x
+#' @export 
+#' @examples
+#' transform_log(1)
+#' transform_log(c(1, 2, 3, 4, 5))
+#' 
+transform_log <- function(x) { 
+  if (!is.numeric(x)) {
+    warning("transform_log: Expecting numeric argument")
+  }
+  x_badval <- is.na(suppressWarnings(as.numeric(x)))
+  x[x_badval] <- 1
+  y <- log(as.numeric(x))
+  y[x_badval] <- NA
+  y
+} 

R/windsorize.R

@@ -1,10 +1,30 @@
 #' Windsorize
 #'
-#' Do some windsorization.
+#' set all outliers to a specified percentile of the data; 
+#' a 90% winsorization would see all data below the 5th percentile set 
+#' to the 5th percentile, 
+#' and data above the 95th percentile set to the 95th percentile.
+#' 
+#' @param x A vector.
+#' @param p A quantile.
+#' @return dataset with trimmed outliers with 10% percentile
+#' @examples
+#' windsorize(c(3,4,4,3,4,5,1))
 #' @export
 windsorize <- function(x, p = .90) {
-  q <- quantile(x, p)
-  x[x >= q] <- q
+  if (length(x) == 0) stop("argument should not be a empty vector")
+  if (all(is.na(x))) {
+    stop("argument should not be a vector containing only NA")
+  }
+  if (!is.numeric(x)) stop("argument should be a numeric vector")
+  if (!is.numeric(p)) stop("argument should be a number from 0 to 1")
+  if (p < 0 || p > 1) {
+    stop("p invalid percentale. Expected values from 0 to 1")
+  }
+  q_lower <- quantile(x, (1-p)/2)
+  q_upper <- quantile(x, 1 - (1-p)/2)
+  x[x <= q_lower] <- q_lower
+  x[x >= q_upper] <- q_upper
   x
 }
 

tests/testthat.R

@@ -0,0 +1,4 @@
+library(testthat)
+library(datacleaner)
+
+test_check("datacleaner")

tests/testthat/test_tranform_log.R

@@ -0,0 +1,11 @@
+context("tranform_log")
+library(datacleaner)
+test_that("tranform_log is correct", {
+  expect_equal(as.character(transform_log(c(2, 2, 2, 2, 3))), as.character(c(0.693147180559945, 0.693147180559945, 0.693147180559945, 0.693147180559945, 1.09861228866811)))
+  expect_equal(as.character(transform_log(c(2, 3, NA))), as.character(c(0.693147180559945, 1.09861228866811, NA)))
+})
+
+test_that("unexpected parameters", {
+  expect_warning(t <- transform_log(c(2, 3, "NA")), "transform_log: Expecting numeric argument")
+  expect_equal(as.character(t), as.character(c(0.693147180559945, 1.09861228866811, NA)))
+})

tests/testthat/test_windorize.R

@@ -0,0 +1,17 @@
+context("Windsorize")
+library(datacleaner)
+test_that("windorizing is correct", {
+  expect_equal(windsorize(c(2,2,2,2,3),.9), c(2,2,2,2,2.8) )
+  expect_equal(windsorize(c(2,2,2,2,1),.9), c(2,2,2,2,1.2) )
+})
+
+test_that("unexpected parameters", {
+  expect_error(windsorize(c(NA,NA,NA), .9), "argument should not be a vector containing only NA")
+  expect_error(windsorize(c(), .9), "argument should not be a empty vector")
+  expect_error(windsorize(c(1,2,3,"4"), .9), "argument should be a numeric vector")
+  expect_error(windsorize(c(1,2,3,4), ".9"), "argument should be a number from 0 to 1")
+  expect_error(windsorize(c(1,2,3,4), -1), "p invalid percentale. Expected values from 0 to 1")
+  expect_error(windsorize(c(1,2,3,4), 1.2), "p invalid percentale. Expected values from 0 to 1")
+  expect_error(windsorize(c(1,2,3,NA), .9))
+
+})