From 80db2afaba0ae346dd150a4c7a066701e146f398 Mon Sep 17 00:00:00 2001
From: "lisa.pizzo" <lisa.pizzo@noreply.imsc.uni-graz.at>
Date: Thu, 4 Dec 2025 11:49:23 +0100
Subject: [PATCH] Main file - Exercise 4

---
 Sheet5/mainEx4.cpp | 168 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 168 insertions(+)
 create mode 100644 Sheet5/mainEx4.cpp

diff --git a/Sheet5/mainEx4.cpp b/Sheet5/mainEx4.cpp
new file mode 100644
index 0000000..0931a35
--- /dev/null
+++ b/Sheet5/mainEx4.cpp
@@ -0,0 +1,168 @@
+// HOW TO COMPILE ON MAC 
+// export CPPFLAGS="-I/opt/homebrew/opt/libomp/include"
+// export LDFLAGS="-L/opt/homebrew/opt/libomp/lib"
+// clang++ -std=c++17 -O3 -Xpreprocessor -fopenmp $CPPFLAGS main.cpp bench_funcs.cpp $LDFLAGS -lomp -o Ex4
+// ./Ex4
+
+
+#include <iostream>
+#include <vector>
+#include <cmath>
+#include <iomanip>
+#include <chrono>
+#include "bench_funcs.h"
+#include <omp.h>   
+
+using namespace std;
+using namespace std::chrono;
+
+void gen_vector_x_y(size_t N, vector<double>& x, vector<double>& y) {
+    x.resize(N);
+    y.resize(N);
+    for (size_t i = 0; i < N; ++i) {
+        x[i] = (i % 219) + 1;
+        y[i] = 1.0 / x[i];
+    }
+}
+
+void gen_matrix_A(size_t M, size_t N, vector<double>& A) {
+    A.resize(M * N);
+    for (size_t i = 0; i < M; ++i)
+        for (size_t j = 0; j < N; ++j)
+            A[i * N + j] = ((i + j) % 219) + 1;
+}
+
+high_resolution_clock::time_point tic_timer;
+void tic() { tic_timer = high_resolution_clock::now(); }
+double toc() {
+    auto t1 = high_resolution_clock::now();
+    duration<double> elapsed = t1 - tic_timer;
+    return elapsed.count();
+}
+
+// CHANGE THE flag VARIABLE IN main() TO CHOOSE A, B, C, D.
+int main() {
+    size_t N = 1'000'000;   
+    size_t M = 3000, L = 3000;
+
+    // Check OpenMP threads
+    #pragma omp parallel
+    {
+        #pragma omp single
+        cout << "Using " << omp_get_num_threads() << " OpenMP threads\n";
+    }
+
+    // Show menu
+    cout << "Choose an option:\n";
+    cout << "  1) SUM + DOT PRODUCT\n";
+    cout << "  2) MATRIX-VECTOR PRODUCT\n";
+    cout << "  3) MATRIX-MATRIX PRODUCT\n";
+    cout << "  4) POLYNOMIAL EVALUATION\n";
+    cout << "Enter your choice (1-4): ";
+
+    int flag;
+    cin >> flag;
+
+    // validate input
+    if(flag < 1 || flag > 4){
+        cout << "Invalid choice. Exiting.\n";
+        return 1;
+    }
+
+    // A) SUM + DOT PRODUCT (parallel)
+    if (flag == 1) {
+        vector<double> x, y;
+        gen_vector_x_y(N, x, y);
+
+        cout << "Running parallel SUM" << endl;
+        tic();
+        double ssum = sum_basic(x); (void)ssum;
+        double dt_sum = toc();
+        cout << "  sum time = " << dt_sum << " s\n";
+
+        cout << "Running parallel DOT PRODUCT" << endl;
+        tic();
+        double sdot = dot_basic(x, y); (void)sdot;
+        double dt_dot = toc();
+        cout << "  dot time = " << dt_dot << " s\n";
+
+        double flops_dot = 2.0 * N;
+        double gflops_dot = (flops_dot / dt_dot) / 1e9;
+
+        cout << "A: N=" << N << "\n";
+        cout << "   SUM time=" << dt_sum << " s\n";
+        cout << "   DOT time=" << dt_dot << " s  GFLOPS=" << gflops_dot << "\n";
+    }
+
+    // B) MATRIX–VECTOR PRODUCT (parallel)
+    else if (flag == 2) {
+        size_t m = M, n = 5000;
+        vector<double> A, x, b;
+
+        gen_matrix_A(m, n, A);
+        x.resize(n);
+        for (size_t j = 0; j < n; ++j)
+            x[j] = 1.0 / (((17 + j) % 219) + 1);
+
+        cout << "Running Matrix times vector (parallel)\n";
+        tic();
+        matvec_rowmajor(A, m, n, x, b);
+        double dt = toc();
+
+        double flops = 2.0 * m * n;
+        double gflops = (flops / dt) / 1e9;
+
+        cout << "B: M=" << m << " N=" << n
+             << " time=" << dt << " s  GFLOPS=" << gflops << endl;
+    }
+
+    // C) MATRIX–MATRIX PRODUCT (parallel)
+    else if (flag == 3) {
+        size_t m = M, l = L, n = 500;
+        vector<double> A, B, C;
+
+        gen_matrix_A(m, l, A);
+        gen_matrix_A(l, n, B);
+
+        cout << "Running Multiplication of matrices (parallel)\n";
+        tic();
+        matmul_rowmajor(A, m, l, B, n, C);
+        double dt = toc();
+
+        double flops = 2.0 * m * l * n;
+        double gflops = (flops / dt) / 1e9;
+
+        cout << "C: M=" << m << " L=" << l << " N=" << n
+             << " time=" << dt << " s  GFLOPS=" << gflops << endl;
+    }
+
+    // D) POLYNOMIAL EVALUATION (parallel Horner)
+    else if (flag == 4) {
+        size_t p = 100; // degree
+        vector<double> a(p+1), x(N), y;
+
+        for (size_t k = 0; k <= p; ++k)
+            a[k] = 1.0 / (k+1);
+
+        for (size_t i = 0; i < N; ++i)
+            x[i] = (i % 219) * 0.001 + 1.0;
+
+        cout << "Running polynomial function (parallel)\n";
+        tic();
+        polyp_horner(a, x, y);
+        double dt = toc();
+
+        double flops = 2.0 * p * N;
+        double gflops = (flops / dt) / 1e9;
+
+        cout << "D: p=" << p << " N=" << N
+             << " time=" << dt << " s  GFLOPS=" << gflops << endl;
+    }
+
+    else {
+        cout << "Invalid flag — choose 1, 2, 3, or 4.\n";
+    }
+
+    cout << "\nDone \n";
+    return 0;
+}