Merge pull request #270 from hosseinmoein/Hossein/Cpp23

Using C++23 to reimplement sort by using zip

Author: hosseinmoein

CMakeLists.txt

@@ -34,7 +34,7 @@ add_library(DataFrame::DataFrame ALIAS DataFrame)
 
 target_sources(DataFrame PRIVATE src/Utils/DateTime.cc)
 
-target_compile_features(DataFrame PUBLIC cxx_std_20)
+target_compile_features(DataFrame PUBLIC cxx_std_23)
 target_compile_definitions(
     DataFrame
     PRIVATE $<$<BOOL:${HMDF_HAVE_CLOCK_GETTIME}>:HMDF_HAVE_CLOCK_GETTIME>

README.md

@@ -24,15 +24,13 @@ ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 -->
+[![C++23](https://img.shields.io/badge/C%2B%2B-23-blue.svg)](https://isocpp.org/std/the-standard )
 [![Build status](https://ci.appveyor.com/api/projects/status/hjw01qui3bvxs8yi?svg=true)](https://ci.appveyor.com/project/hosseinmoein/dataframe)
+<BR>
 ![GitHub](https://img.shields.io/github/license/hosseinmoein/DataFrame.svg?color=red&style=popout)
-[![C++20](https://img.shields.io/badge/C%2B%2B-20-blue.svg)](https://isocpp.org/std/the-standard )
 [![Codacy Badge](https://api.codacy.com/project/badge/Grade/db646376a4014c3788c7224e670fe451)](https://app.codacy.com/manual/hosseinmoein/DataFrame?utm_source=github.com&utm_medium=referral&utm_content=hosseinmoein/DataFrame&utm_campaign=Badge_Grade_Dashboard)
 <BR>
-[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/hosseinmoein/DataFrame/master)
-[![Maintenance](https://img.shields.io/badge/Maintained%3F-yes-green.svg)](https://GitHub.com/hosseinmoein/DataFrame/graphs/commit-activity)
 ![GitHub tag (latest by date)](https://img.shields.io/github/tag-date/hosseinmoein/DataFrame.svg?color=blue&label=Official%20Release&style=popout)
-<BR>
 ![Conan Center](https://img.shields.io/conan/v/dataframe)
 [![VCPKG package](https://repology.org/badge/version-for-repo/vcpkg/dataframe.svg)](https://repology.org/project/dataframe/versions)
 

benchmarks/dataframe_performance.cc

@@ -36,7 +36,8 @@ using namespace hmdf;
 using namespace std::chrono;
 
 constexpr std::size_t   ALIGNMENT = 64;
-constexpr std::size_t   SIZE = 300000000;
+// constexpr std::size_t   SIZE = 300000000;
+constexpr std::size_t   SIZE = 10000000;
 
 typedef StdDataFrame64<time_t> MyDataFrame;
 
@@ -57,7 +58,7 @@ int main(int, char *[]) {
 
     std::cout << "Data generation/load time: "
               << double(duration_cast<microseconds>(second - first).count()) / 1000000.0
-              << std::endl;
+              << " secs" << std::endl;
 
     MeanVisitor<double, time_t> n_mv;
     VarVisitor<double, time_t>  ln_vv;
@@ -81,14 +82,25 @@ int main(int, char *[]) {
 
     const auto  fourth = high_resolution_clock::now();
 
+    // df.sort<double, double, double>("log_normal", sort_spec::ascen,
+    //                                 "exponential", sort_spec::ascen);
+    // std::cout << "1001th value in normal column: "
+    //           << df.get_column<double>("normal")[1001] << std::endl;
+
+    const auto  fifth = high_resolution_clock::now();
+
     std::cout << "Calculation time: "
               << double(duration_cast<microseconds>(third - second).count()) / 1000000.0
-              << '\n'
+              << " secs\n"
               << "Selection time: "
               << double(duration_cast<microseconds>(fourth - third).count()) / 1000000.0
-              << '\n'
+              << " secs\n"
+              // << "Sorting time: "
+              // << double(duration_cast<microseconds>(fifth - fourth).count()) / 1000000.0
+              // << " secs\n"
               << "Overall time: "
-              << double(duration_cast<microseconds>(fourth - first).count()) / 1000000.0
+              << double(duration_cast<microseconds>(fifth - first).count()) / 1000000.0
+              << " secs"
               << std::endl;
     return (0);
 }

benchmarks/polars_performance.py

@@ -4,7 +4,8 @@
 
 # ------------------------------------------------------------------------------
 
-SIZE: int = 300000000
+# SIZE: int = 300000000
+SIZE: int = 10000000
 
 first = datetime.datetime.now()
 df = pl.DataFrame({"normal": np.random.normal(size=SIZE),
@@ -13,7 +14,7 @@
                    })
 second = datetime.datetime.now()
 print(f"Data generation/load time: "
-      f"{(second - first).seconds}.{(second - first).microseconds}")
+      f"{(second - first).seconds}.{(second - first).microseconds} secs")
 
 df2 = df.select(
    mean = pl.col("normal").mean(),
@@ -32,9 +33,14 @@
 print(f"Number of rows after select: {df3.select(pl.count()).item()}")
 fourth = datetime.datetime.now()
 
-print(f"Calculation time: {(third - second).seconds}.{(third - second).microseconds}")
-print(f"Selection time: {(fourth - third).seconds}.{(fourth - third).microseconds}")
-print(f"Overall time: {(fourth - first).seconds}.{(fourth - first).microseconds}")
+# df4 = df.sort(["log_normal", "exponential"]);
+# print(f"1001th value in normal column: {df4['normal'][1001]}")
+fifth = datetime.datetime.now()
+
+print(f"Calculation time: {(third - second).seconds}.{(third - second).microseconds} secs")
+print(f"Selection time: {(fourth - third).seconds}.{(fourth - third).microseconds} secs")
+# print(f"Sorting time: {(fifth - fourth).seconds}.{(fifth - fourth).microseconds} secs")
+print(f"Overall time: {(fifth - first).seconds}.{(fifth - first).microseconds} secs")
 
 # ------------------------------------------------------------------------------
 

include/DataFrame/DataFrameStatsVisitors.h

@@ -4842,6 +4842,9 @@ struct  LowessVisitor  {
                 const Y &y_begin, const Y &y_end,  // dependent variable
                 const X &x_begin, const X &x_end)  {  // independent variable
 
+        using bool_vec_t =
+            std::vector<bool, typename allocator_declare<bool, A>::type>;
+
         assert(frac_ >= 0 && frac_ <= 1);
         assert(loop_n_ > 2);
 
@@ -4862,7 +4865,10 @@ struct  LowessVisitor  {
                       [] (auto lhs, auto rhs) -> bool  {
                           return (lhs < rhs);
                       });
-            _sort_by_sorted_index_(yvals, sorting_idxs, col_s);
+
+            bool_vec_t  done_vec (col_s);
+
+            _sort_by_sorted_index_(yvals, sorting_idxs, done_vec, col_s);
             lowess_(idx_begin, idx_end,
                     yvals.begin(), yvals.end(),
                     xvals.begin(), xvals.end());

include/DataFrame/DataFrameTypes.h

@@ -579,10 +579,10 @@ struct  RandGenParams  {
     std::size_t     t_dist { 1 };
     // The μ distribution parameter (the mean of the distribution)
     //
-    double          mean { 1.0 };
+    double          mean { 0 };
     // the σ distribution parameter (standard deviation)
     //
-    double          std { 0 };
+    double          std { 1 };
     // The λ distribution parameter (the rate parameter)
     //
     double          lambda { 1.0 };