[incubator-datasketches-cpp] branch master updated: cpc union moving update

[al...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

alsay pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-datasketches-cpp.git


The following commit(s) were added to refs/heads/master by this push:
     new 176d74e  cpc union moving update
176d74e is described below

commit 176d74e90a1b85990dc5dd6ed9a09ae097367e8a
Author: AlexanderSaydakov <Al...@users.noreply.github.com>
AuthorDate: Tue Feb 25 14:31:58 2020 -0800

    cpc union moving update
---
 cpc/include/cpc_union.hpp      |  3 +++
 cpc/include/cpc_union_impl.hpp | 15 +++++++++++++--
 cpc/test/cpc_union_test.cpp    | 10 ++++++++++
 3 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/cpc/include/cpc_union.hpp b/cpc/include/cpc_union.hpp
index 9601cc0..9bdae6e 100644
--- a/cpc/include/cpc_union.hpp
+++ b/cpc/include/cpc_union.hpp
@@ -53,6 +53,7 @@ class cpc_union_alloc {
     cpc_union_alloc<A>& operator=(cpc_union_alloc<A>&& other) noexcept;
 
     void update(const cpc_sketch_alloc<A>& sketch);
+    void update(cpc_sketch_alloc<A>&& sketch);
     cpc_sketch_alloc<A> get_result() const;
 
   private:
@@ -65,6 +66,8 @@ class cpc_union_alloc {
     cpc_sketch_alloc<A>* accumulator;
     vector_u64<A> bit_matrix;
 
+    template<typename S> void internal_update(S&& sketch); // to support both rvalue and lvalue
+
     cpc_sketch_alloc<A> get_result_from_accumulator() const;
     cpc_sketch_alloc<A> get_result_from_bit_matrix() const;
 
diff --git a/cpc/include/cpc_union_impl.hpp b/cpc/include/cpc_union_impl.hpp
index 3363b77..bd390dc 100644
--- a/cpc/include/cpc_union_impl.hpp
+++ b/cpc/include/cpc_union_impl.hpp
@@ -85,6 +85,17 @@ cpc_union_alloc<A>& cpc_union_alloc<A>::operator=(cpc_union_alloc<A>&& other) no
 
 template<typename A>
 void cpc_union_alloc<A>::update(const cpc_sketch_alloc<A>& sketch) {
+  internal_update(sketch);
+}
+
+template<typename A>
+void cpc_union_alloc<A>::update(cpc_sketch_alloc<A>&& sketch) {
+  internal_update(std::forward<cpc_sketch_alloc<A>>(sketch));
+}
+
+template<typename A>
+template<typename S>
+void cpc_union_alloc<A>::internal_update(S&& sketch) {
   const uint16_t seed_hash_union = compute_seed_hash(seed);
   const uint16_t seed_hash_sketch = compute_seed_hash(sketch.seed);
   if (seed_hash_union != seed_hash_sketch) {
@@ -106,9 +117,9 @@ void cpc_union_alloc<A>::update(const cpc_sketch_alloc<A>& sketch) {
         cpc_sketch_alloc<A>::flavor::SPARSE != initial_dest_flavor) throw std::logic_error("wrong flavor");
 
     // The following partially fixes the snowplow problem provided that the K's are equal.
-    // A complete fix is coming soon.
     if (cpc_sketch_alloc<A>::flavor::EMPTY == initial_dest_flavor and lg_k == sketch.get_lg_k()) {
-      *accumulator = sketch;
+      *accumulator = std::forward<S>(sketch);
+      return;
     }
 
     walk_table_updating_sketch(sketch.surprising_value_table);
diff --git a/cpc/test/cpc_union_test.cpp b/cpc/test/cpc_union_test.cpp
index 256ee10..a9928a9 100644
--- a/cpc/test/cpc_union_test.cpp
+++ b/cpc/test/cpc_union_test.cpp
@@ -37,6 +37,7 @@ class cpc_union_test: public CppUnit::TestFixture {
   CPPUNIT_TEST(reduce_k_empty);
   CPPUNIT_TEST(reduce_k_sparse);
   CPPUNIT_TEST(reduce_k_window);
+  CPPUNIT_TEST(moving_update);
   CPPUNIT_TEST_SUITE_END();
 
   void lg_k_limits() {
@@ -149,6 +150,15 @@ class cpc_union_test: public CppUnit::TestFixture {
     CPPUNIT_ASSERT_DOUBLES_EQUAL(1000, r.get_estimate(), 1000 * RELATIVE_ERROR_FOR_LG_K_11);
   }
 
+  void moving_update() {
+    cpc_union u(11);
+    cpc_sketch s(11);
+    for (int i = 0; i < 100; i++) s.update(i); // sparse
+    u.update(std::move(s));
+    cpc_sketch r = u.get_result();
+    CPPUNIT_ASSERT_DOUBLES_EQUAL(100, r.get_estimate(), 100 * RELATIVE_ERROR_FOR_LG_K_11);
+  }
+
 };
 
 CPPUNIT_TEST_SUITE_REGISTRATION(cpc_union_test);


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@datasketches.apache.org
For additional commands, e-mail: commits-help@datasketches.apache.org