Merge pull request #42 from r-spatial/fix_box

Fix box

Author: rsbivand

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: classInt
-Version: 0.4-8
-Date: 2022-09-22
+Version: 0.4-9
+Date: 2023-02-21
 Title: Choose Univariate Class Intervals
 Authors@R: c(
   person("Roger", "Bivand", role=c("aut", "cre"), email="[email protected]", comment=c(ORCID="0000-0003-2392-6140")),
@@ -16,7 +16,8 @@ Suggests:
     spData (>= 0.2.6.2),
     units,
     knitr,
-    rmarkdown
+    rmarkdown,
+    tinytest
 NeedsCompilation: yes
 Description: Selected commonly used methods for choosing univariate class intervals for mapping or other graphics purposes.
 License: GPL (>= 2)

NEWS.md

@@ -1,3 +1,7 @@
+## Version 0.4-9
+
+- #41 issues. The maximum and minimum breaks are set to \code{+Inf} and \code{-Inf} to avoid errors induced in the earlier version where breaks could cease to be strictly ascending. The \code{legacy=} argument with value \code{TRUE} may be used to revert to the previous behaviour.
+
 ## Version 0.4-8
 
 - #18 and #38: `classIntervals()` has a new style `"box"`, where a box map is an augmented quartile map, with an additional lower and upper category. When there are lower outliers, then the starting point for the breaks is the minimum value, and the second break is the lower fence. When there are no lower outliers, then the starting point for the breaks will be the lower fence, and the second break is the minimum value (there will be no observations that fall in the interval between the lower fence and the minimum value) (@angela-li, @dieghernan).

R/classInt.R

@@ -357,7 +357,9 @@ classIntervals <- function(var, n, style="quantile", rtimes=3, ..., intervalClos
 
         dots <- list(...)
         iqr_mult <- ifelse(is.null(dots$iqr_mult), 1.5, dots$iqr_mult)
+        stopifnot(iqr_mult >= 0)
         qtype <- ifelse(is.null(dots$type), 7, dots$type)
+        legacy <- ifelse(is.null(dots$legacy), FALSE, dots$legacy)
 
         qv <- unname(quantile(var, type=qtype))
         iqr <- iqr_mult * (qv[4] - qv[2])
@@ -369,15 +371,25 @@ classIntervals <- function(var, n, style="quantile", rtimes=3, ..., intervalClos
 
         # logic for lower and upper fences
         if (lofence < qv[1]) {  # no lower outliers
-          bb[1] <- lofence
-          bb[2] <- floor(qv[1])
+          if (legacy) {
+            bb[1] <- lofence
+            bb[2] <- floor(qv[1])
+          } else {
+            bb[1] <- -Inf
+            bb[2] <- lofence
+          }
         } else {
           bb[2] <- lofence
           bb[1] <- qv[1]
         }
         if (upfence > qv[5]) { # no upper outliers
-          bb[7] <- upfence
-          bb[6] <- ceiling(qv[5])
+          if (legacy) {
+            bb[7] <- upfence
+            bb[6] <- ceiling(qv[5])
+          } else {
+            bb[7] <- +Inf
+            bb[6] <- upfence
+          }
         } else {
           bb[6] <- upfence
           bb[7] <- qv[5]

inst/tinytest/test_box.R

@@ -0,0 +1,6 @@
+library(classInt)
+set.seed(101)
+x <- rnorm(50)
+expect_error(print(classIntervals(x, style="box", iqr_mult=1.5, type=5, legacy=TRUE)))
+expect_inherits(classIntervals(x, style="box", iqr_mult=1.5, type=5), "classIntervals")
+expect_error(classIntervals(x, style="box", iqr_mult=-1))

man/classIntervals.Rd

@@ -70,7 +70,7 @@ classIntervals2shingle(x)
 
   The "maximum" style uses the Maximum Breaks method of classification finding the k - 1 largest differences in \code{var}. The mean of the values that generated the largest splits is used as the interval boundary.
   
-  The "box" style generates 7 breaks (therefore 6 categories) based on a box-and-whisker plot. First and last categories include the data values considered as outliers, and the four remaining categories are defined by the percentiles 25, 50 and 75 of the data distribution. By default, the identification of outliers is based on the interquantile range (IQR), so values lower than percentile 25 - 1.5 * IQR or higher than percentile 75 + 1.5 * IQR are considered as outliers. The multiplier applied to the IQR \code{iqr_mult = 1.5} may be modified through \code{\dots}. As in the \code{"quantile"} style, the \code{type=} argument may be used to choose the quantile algoritm (default 7).
+  The "box" style generates 7 breaks (therefore 6 categories) based on a box-and-whisker plot. First and last categories include the data values considered as outliers, and the four remaining categories are defined by the percentiles 25, 50 and 75 of the data distribution. By default, the identification of outliers is based on the interquantile range (IQR), so values lower than percentile 25 - 1.5 * IQR or higher than percentile 75 + 1.5 * IQR are considered as outliers. The multiplier applied to the IQR \code{iqr_mult = 1.5} may be modified through \code{\dots}; the value must not be negative. As in the \code{"quantile"} style, the \code{type=} argument may be used to choose the quantile algoritm (default 7, standard boxplots use 5 or 2). From 0.4-9 and #41, the maximum and minimum are set to \code{+Inf} and \code{-Inf} to avoid errors induced in the earlier version where breaks could cease to be strictly ascending. The \code{legacy=} argument with value \code{TRUE} may be used to revert to the previous behaviour.
 
 }
 

tests/tinytest.R

@@ -0,0 +1,3 @@
+if ( requireNamespace("tinytest", quietly=TRUE) ){
+  tinytest::test_package("classInt")
+}