expressions: Implement the MEDIAN function.

diff --git a/NEWS b/NEWS
index 4236ae1754d12df1e8fd02e2164a6e40e9329e0a..309cf9e2a5a1a4915d0e1fc13c504023ab0a9c64 100644 (file)
--- a/NEWS
+++ b/NEWS
@@ -38,6 +38,12 @@ Changes from 0.8.5 to 0.9.0:
    discourage its use PSPP and PSPPIRE do not directly read or write
    this format.
 
+ * The following functions for transformation expressions are new:
+
+   - REPLACE, for search-and-replace of one string with another.
+
+   - MEDIAN, to compute the median of its arguments.
+
  * Bug fixes, including the following notable ones:
 
    - The correlation coefficient in the paired samples t-test

diff --git a/doc/expressions.texi b/doc/expressions.texi
index b7e23a5f96c3d0721a4614832a09292a47bd67e9..e468b8750ac457361b878a7788ac24e6d0e8a544 100644 (file)
--- a/doc/expressions.texi
+++ b/doc/expressions.texi
@@ -262,7 +262,7 @@ The sections below describe each function in detail.
 * Trigonometry::                ACOS ARCOS ARSIN ARTAN ASIN ATAN COS SIN TAN
 * Missing Value Functions::     MISSING NMISS NVALID SYSMIS VALUE
 * Set Membership::              ANY RANGE
-* Statistical Functions::       CFVAR MAX MEAN MIN SD SUM VARIANCE
+* Statistical Functions::       CFVAR MAX MEAN MEDIAN MIN SD SUM VARIANCE
 * String Functions::            CONCAT INDEX LENGTH LOWER LPAD LTRIM NUMBER 
                                 REPLACE RINDEX RPAD RTRIM STRING SUBSTR UPCASE
 * Time and Date::               CTIME.xxx DATE.xxx TIME.xxx XDATE.xxx
@@ -510,6 +510,13 @@ be numeric or string.
 Results in the mean of the values of @var{number}.
 @end deftypefn
 
+@cindex median
+@deftypefn {Function} {} MEDIAN (@var{number}, @var{number}[, @dots{}])
+Results in the median of the values of @var{number}.  Given an even
+number of nonmissing arguments, yields the mean of the two middle
+values.
+@end deftypefn
+
 @cindex minimum
 @deftypefn {Function} {} MIN (@var{number}, @var{number}[, @dots{}])
 Results in the value of the least @var{value}.  The @var{value}s may

diff --git a/src/language/expressions/helpers.c b/src/language/expressions/helpers.c
index cfc46290b785cfad8c4d5dd8182c65fe016774cd..57e17713b994f9036c97ddda0d829035fb726605 100644 (file)
--- a/src/language/expressions/helpers.c
+++ b/src/language/expressions/helpers.c
@@ -717,3 +717,32 @@ replace_string (struct expression *e,
 
   return result;
 }
+
+static int
+compare_doubles (const void *a_, const void *b_)
+{
+  const double *ap = a_;
+  const double *bp = b_;
+  double a = *ap;
+  double b = *bp;
+
+  /* Sort SYSMIS to the end. */
+  return (a == b ? 0
+          : a == SYSMIS ? 1
+          : b == SYSMIS ? -1
+          : a > b ? 1 : -1);
+}
+
+double
+median (double *a, size_t n)
+{
+  /* Sort the array in-place, sorting SYSMIS to the end. */
+  qsort (a, n, sizeof *a, compare_doubles);
+
+  /* Drop SYSMIS. */
+  n = count_valid (a, n);
+
+  return (!n ? SYSMIS
+          : n % 2 ? a[n / 2]
+          : (a[n / 2 - 1] + a[n / 2]) / 2.0);
+}

diff --git a/src/language/expressions/helpers.h b/src/language/expressions/helpers.h
index 51e1db0329a1bf3d8cd7c83f72e3023aadf5b6a1..c4e332340625f17d9f52962b743eaa00c11efb9e 100644 (file)
--- a/src/language/expressions/helpers.h
+++ b/src/language/expressions/helpers.h
@@ -89,4 +89,6 @@ struct substring replace_string (struct expression *,
                                  struct substring replacement,
                                  double n);
 
+double median (double *, size_t n);
+
 #endif /* expressions/helpers.h */

diff --git a/src/language/expressions/operations.def b/src/language/expressions/operations.def
index 203ccbb42a2b9825135d624c6700bc67b73e0cee..18ab4447c824032a7f896161c9fc760fa31a2d62 100644 (file)
--- a/src/language/expressions/operations.def
+++ b/src/language/expressions/operations.def
@@ -168,6 +168,11 @@ function MEAN.1 (a[n])
   return mean;
 }
 
+function MEDIAN.1 (a[n])
+{
+  return median (a, n);
+}
+
 function MIN.1 (a[n])
 {
   double min;

diff --git a/tests/language/expressions/evaluate.at b/tests/language/expressions/evaluate.at
index 41104dd70b422d8287da58c604647a10d21ca340..703758b3b598d9602ecf0f35ec2bfaa1515dbcd3 100644 (file)
--- a/tests/language/expressions/evaluate.at
+++ b/tests/language/expressions/evaluate.at
@@ -633,7 +633,7 @@ MIN(string[, string]...).]],
   [[min("1", "2")], ["1"]],
   [[min("1")], ["1"]])
 
-CHECK_EXPR_EVAL([cfvar mean sd sum variance],
+CHECK_EXPR_EVAL([cfvar mean median sd sum variance],
   [[cfvar(1, 2, 3, 4, 5)], [0.53]],
   [[cfvar(1, $sysmis, 2, 3, $sysmis, 4, 5)], [0.53]],
   [[cfvar(1, 2)], [0.47]],
@@ -661,6 +661,25 @@ CHECK_EXPR_EVAL([cfvar mean sd sum variance],
   [[mean.4(1, 2, 3)], [error],
    [error: DEBUG EVALUATE: With MEAN(number[, number]...), using minimum valid argument count of 4 does not make sense when passing only 3 arguments in list.]],
 
+  [[median(1, 2, 3, 4, 5)], [3.00]],
+  [[median(2, 3, 4, 5, 1)], [3.00]],
+  [[median(2, 3, 4, 1, 5)], [3.00]],
+  [[median(2, 1, 4, 5, 3)], [3.00]],
+  [[median(1, 2, 3, 4)], [2.50]],
+  [[median(2, 3, 1, 4)], [2.50]],
+  [[median(2, 3, 4, 1)], [2.50]],
+  [[median(2, 1, 4, 3)], [2.50]],
+  [[median(1, $sysmis, 3, 4, 5)], [3.50]],
+  [[median(2, 3, 4, 5, $sysmis, 1)], [3.00]],
+  [[median($sysmis, $sysmis, $sysmis, 2, 3, 4, 1, 5)], [3.00]],
+  [[median(1, 2, 3)], [2.00]],
+  [[median(1)], [1.00]],
+  [[median(1, 2)], [1.50]],
+  [[median(1, 2, $sysmis)], [1.50]],
+  [[median(1, $sysmis, $sysmis)], [1.00]],
+  [[median($sysmis, $sysmis, $sysmis)], [sysmis]],
+  [[median.3(1, 2, $sysmis)], [sysmis]],
+  [[median.2(1, $sysmis)], [sysmis]],
 
   [[sd(1, 2, 3, 4, 5)], [1.58]],
   [[sd(1, $sysmis, 2, 3, $sysmis, 4, 5)], [1.58]],