enumerable-statistics 2.0.3 → 2.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 85d1034d8bb8062a6d2469ab813518af28acad6b3a4e47e04335e4509f17f830
-  data.tar.gz: 9886d7f3e791960f499666c4925263a9c163129a9cb8957907ac431d837d5c71
+  metadata.gz: 3fa86766eb3c5369422d08bac77705368b0bde5faa216eecd9f78115eb7548b6
+  data.tar.gz: 313dc69fe5d6b193278489b576af4bb90679eebf6814c36a2924a1241cd5c8be
 SHA512:
-  metadata.gz: 5ccdb91def90bb4740d65bb01bd67fe8bae3932f4e1bb61180862e0b509ab92f7efe4034fcb9855ae0314e009d093c1f2852fe9f3126733f29b68fa706f60a31
-  data.tar.gz: 59cffac4f18c6dae5dee8da57c222ce1fbb14594ef3e6e43c53d28bf1f606512c4c7f75a585d8b529992df2eb0c173bbd355b5e7b505d63ef0befdf1168aff6c
+  metadata.gz: 13fd8a73ae4e83aa85210cbbec1882a2129802aa377f7100ffa72e48ec7df685ba5eff1553cd2ef0b29eddd2b13715769de822adcc881c6d57122212d4c8c13a
+  data.tar.gz: d4dc08f32d2d9a43d743e213b2725f0bc41ac7d6d14c68ccecd1b3563d022c8c1c8b60eea682bce554bc7b5425e3c4b552766bb8b3a7faa386164ee8f00303b7

data/CHANGELOG.md

@@ -1,3 +1,12 @@
+# 2.0.5
+
+- Add weighted histogram support
+
+# 2.0.4
+
+- Add `find_min`, `find_max`, `argmin`, `argmax` methods
+- Fix `nbin=:auto` case in `histogram` method
+
 # 2.0.3
 
 - Ractor-safe version

data/ext/enumerable/statistics/extension/array_ext.c

@@ -0,0 +1,60 @@
+#include <ruby/ruby.h>
+
+static VALUE
+ary_find_max(VALUE ary)
+{
+  const long n = RARRAY_LEN(ary);
+  if (n == 0) {
+    return Qnil;
+  }
+
+  long imax = 0;
+  VALUE max = RARRAY_AREF(ary, imax);
+
+  long i;
+  for (i = 1; i < n; ++i) {
+    VALUE v = RARRAY_AREF(ary, i);
+    if (RTEST(rb_funcall(v, '>', 1, max))) {
+      imax = i;
+      max = v;
+    }
+  }
+
+  return rb_assoc_new(max, LONG2NUM(imax));
+}
+
+static VALUE
+ary_find_min(VALUE ary)
+{
+  const long n = RARRAY_LEN(ary);
+  if (n == 0) {
+    return Qnil;
+  }
+
+  long imin = 0;
+  VALUE min = RARRAY_AREF(ary, imin);
+
+  long i;
+  for (i = 1; i < n; ++i) {
+    VALUE v = RARRAY_AREF(ary, i);
+    if (RTEST(rb_funcall(v, '<', 1, min))) {
+      imin = i;
+      min = v;
+    }
+  }
+
+  return rb_assoc_new(min, LONG2NUM(imin));
+}
+
+void
+Init_array_extension(void)
+{
+  VALUE mEnumerableStatistics = rb_const_get_at(rb_cObject, rb_intern("EnumerableStatistics"));
+  VALUE mArrayExtension = rb_const_get_at(mEnumerableStatistics, rb_intern("ArrayExtension"));
+
+  rb_undef_method(mArrayExtension, "find_max");
+  rb_define_method(mArrayExtension, "find_max", ary_find_max, 0);
+
+  rb_undef_method(mArrayExtension, "find_min");
+  rb_define_method(mArrayExtension, "find_min", ary_find_min, 0);
+}

data/ext/enumerable/statistics/extension/statistics.c

@@ -97,7 +97,7 @@ static ID idPow, idPLUS, idMINUS, idSTAR, idDIV, idGE;
 static ID id_eqeq_p, id_idiv, id_negate, id_to_f, id_cmp, id_nan_p;
 static ID id_each, id_real_p, id_sum, id_population, id_closed, id_edge;
 
-static VALUE sym_left, sym_right;
+static VALUE sym_auto, sym_left, sym_right;
 
 static VALUE cHistogram;
 
@@ -2102,50 +2102,61 @@ histogram_edge_bin_index(VALUE edge, VALUE rb_x, int left_p)
 }
 
 static void
-histogram_weights_push_values(VALUE weights, VALUE edge, VALUE values, int left_p)
+histogram_weights_push_values(VALUE bin_weights, VALUE edge, VALUE values, VALUE weight_array, int left_p)
 {
   VALUE x, cur;
-  long i, n, bi;
+  long i, n, bi, one, weighted = 0;
 
   n = RARRAY_LEN(values);
+
+  if (! NIL_P(weight_array)) {
+    assert(RB_TYPE_P(weight_array, T_ARRAY));
+    assert(RARRAY_LEN(weight_array) == n);
+    weighted = 1;
+  }
+
+  one = INT2FIX(1);
   for (i = 0; i < n; ++i) {
     x = RARRAY_AREF(values, i);
 
+    VALUE w;
+    if (weighted) {
+      w = RARRAY_AREF(weight_array, i);
+      if (RB_TYPE_P(w, T_COMPLEX)) {
+        VALUE imag = RCOMPLEX(w)->imag;
+        if (! RTEST(f_zero_p(imag))) {
+          goto type_error;
+        }
+      }
+      else if (rb_obj_is_kind_of(w, rb_cNumeric)) {
+        if (!RTEST(f_real_p(w))) {
+          goto type_error;
+        }
+      }
+      else {
+        goto type_error;
+      }
+    }
+    else {
+      w = one;
+    }
+
     bi = histogram_edge_bin_index(edge, x, left_p);
 
-    cur = rb_ary_entry(weights, bi);
+    cur = rb_ary_entry(bin_weights, bi);
     if (NIL_P(cur)) {
-      cur = INT2FIX(1);
+      cur = w;
     }
     else {
-      cur = rb_funcall(cur, idPLUS, 1, INT2FIX(1));
+      cur = rb_funcall(cur, idPLUS, 1, w);
     }
 
-    rb_ary_store(weights, bi, cur);
-  }
-}
-
-static int
-opt_closed_left_p(VALUE opts)
-{
-  int left_p = 1;
-
-  if (!NIL_P(opts)) {
-    VALUE closed;
-#ifdef HAVE_RB_GET_KWARGS
-    ID kwargs = id_closed;
-    rb_get_kwargs(opts, &kwargs, 0, 1, &closed);
-#else
-    closed = rb_hash_lookup2(opts, ID2SYM(id_closed), sym_left);
-#endif
-    left_p = (closed != sym_right);
-    if (left_p && closed != sym_left) {
-      rb_raise(rb_eArgError, "invalid value for :closed keyword "
-               "(%"PRIsVALUE" for :left or :right)", closed);
-    }
+    rb_ary_store(bin_weights, bi, cur);
   }
+  return;
 
-  return left_p;
+type_error:
+      rb_raise(rb_eTypeError, "weight array must have only real numbers");
 }
 
 static inline long
@@ -2270,9 +2281,12 @@ ary_histogram_calculate_edge(VALUE ary, const long nbins, const int left_p)
 }
 
 /* call-seq:
- *    ary.histogram(nbins=:auto, closed: :left)
+ *    ary.histogram(nbins=:auto, weight: nil, closed: :left)
  *
  * @param [Integer] nbins  The approximate number of bins
+ * @params [Array<Numeric>] weight
+ *   An optional weight array, that has the same length as the receiver.
+ *   `weight[i]` means the weight value of the i-th element in the receiver.
  * @param [:left, :right] closed
  *   If :left (the default), the bin interval are left-closed.
  *   If :right, the bin interval are right-closed.
@@ -2282,30 +2296,63 @@ ary_histogram_calculate_edge(VALUE ary, const long nbins, const int left_p)
 static VALUE
 ary_histogram(int argc, VALUE *argv, VALUE ary)
 {
-  VALUE arg0, opts, edge, weights;
-  int left_p;
-  long nbins, nweights, i;
+  VALUE arg0, kwargs, edge, bin_weights;
+  long nbins, n_bin_weights, i;
 
-  rb_scan_args(argc, argv, "01:", &arg0, &opts);
-  if (NIL_P(arg0)) {
+  VALUE weight_array = Qnil;
+  int left_p = 1;
+
+  rb_scan_args(argc, argv, "01:", &arg0, &kwargs);
+  if (NIL_P(arg0) || arg0 == sym_auto) {
     nbins = sturges(RARRAY_LEN(ary));
   }
   else {
     nbins = NUM2LONG(arg0);
   }
-  left_p = opt_closed_left_p(opts);
+
+  if (!NIL_P(kwargs)) {
+    enum { kw_weight, kw_closed };
+    static ID kwarg_keys[2];
+    VALUE kwarg_vals[2];
+    VALUE closed;
+
+    if (!kwarg_keys[0]) {
+      kwarg_keys[kw_weight] = rb_intern("weight");
+      kwarg_keys[kw_closed] = rb_intern("closed");
+    }
+
+    rb_get_kwargs(kwargs, kwarg_keys, 0, 2, kwarg_vals);
+
+    weight_array = kwarg_vals[kw_weight];
+    if (weight_array != Qundef) {
+      weight_array = rb_check_convert_type(weight_array, T_ARRAY, "Array", "to_ary");
+      if (RARRAY_LEN(weight_array) != RARRAY_LEN(ary)) {
+        rb_raise(rb_eArgError, "weight array must have the same number of items as the receiver array");
+      }
+    }
+    else {
+      weight_array = Qnil;
+    }
+
+    closed = kwarg_vals[kw_closed];
+    left_p = (closed != sym_right);
+    if (left_p && closed != Qundef && closed != sym_left) {
+      rb_raise(rb_eArgError, "invalid value for :closed keyword "
+               "(%"PRIsVALUE" for :left or :right)", closed);
+    }
+  }
 
   edge = ary_histogram_calculate_edge(ary, nbins, left_p);
 
-  nweights = RARRAY_LEN(edge) - 1;
-  weights = rb_ary_new_capa(nweights);
-  for (i = 0; i < nweights; ++i) {
-    rb_ary_store(weights, i, INT2FIX(0));
+  n_bin_weights = RARRAY_LEN(edge) - 1;
+  bin_weights = rb_ary_new_capa(n_bin_weights);
+  for (i = 0; i < n_bin_weights; ++i) {
+    rb_ary_store(bin_weights, i, INT2FIX(0));
   }
 
-  histogram_weights_push_values(weights, edge, ary, left_p);
+  histogram_weights_push_values(bin_weights, edge, ary, weight_array, left_p);
 
-  return rb_struct_new(cHistogram, edge, weights,
+  return rb_struct_new(cHistogram, edge, bin_weights,
                        left_p ? sym_left : sym_right,
                        Qfalse);
 }
@@ -2352,6 +2399,9 @@ Init_extension(void)
 
   rb_define_method(rb_cArray, "histogram", ary_histogram, -1);
 
+  void Init_array_extension(void);
+  Init_array_extension();
+
   idPLUS = '+';
   idMINUS = '-';
   idSTAR = '*';
@@ -2371,6 +2421,7 @@ Init_extension(void)
   id_closed = rb_intern("closed");
   id_edge = rb_intern("edge");
 
+  sym_auto = ID2SYM(rb_intern("auto"));
   sym_left = ID2SYM(rb_intern("left"));
   sym_right = ID2SYM(rb_intern("right"));
 }

data/lib/enumerable_statistics.rb

@@ -1,2 +1,3 @@
 require_relative "enumerable_statistics/version"
+require_relative "enumerable_statistics/array_ext"
 require_relative "enumerable_statistics/histogram"

data/lib/enumerable_statistics/array_ext.rb

@@ -0,0 +1,37 @@
+module EnumerableStatistics
+  module ArrayExtension
+    def find_max
+      n = size
+      return nil if n == 0
+
+      imax, i = 0, 1
+      while i < n
+        imax = i if self[i] > self[imax]
+        i += 1
+      end
+      [self[imax], imax]
+    end
+
+    def argmax
+      find_max[1]
+    end
+
+    def find_min
+      n = size
+      return nil if n == 0
+
+      imin, i = 0, 1
+      while i < n
+        imin = i if self[i] < self[imax]
+        i += 1
+      end
+      [self[imin], imin]
+    end
+
+    def argmin
+      find_min[1]
+    end
+  end
+
+  Array.include ArrayExtension
+end

data/lib/enumerable_statistics/version.rb

@@ -1,5 +1,5 @@
 module EnumerableStatistics
-  VERSION = '2.0.3'
+  VERSION = '2.0.5'
 
   module Version
     numbers, TAG = VERSION.split('-', 2)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: enumerable-statistics
 version: !ruby/object:Gem::Version
-  version: 2.0.3
+  version: 2.0.5
 platform: ruby
 authors:
 - Kenta Murata
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-01-04 00:00:00.000000000 Z
+date: 2021-06-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -149,11 +149,13 @@ files:
 - bin/rspec
 - bin/setup
 - enumerable-statistics.gemspec
+- ext/enumerable/statistics/extension/array_ext.c
 - ext/enumerable/statistics/extension/extconf.rb
 - ext/enumerable/statistics/extension/statistics.c
 - images/benchmark.png
 - lib/enumerable/statistics.rb
 - lib/enumerable_statistics.rb
+- lib/enumerable_statistics/array_ext.rb
 - lib/enumerable_statistics/histogram.rb
 - lib/enumerable_statistics/version.rb
 - templates/default/layout/html/headers.erb