✨ Parallelised the monte carlo code for pi calculation.

answers/01/pi_monte_carlo_parallel.cpp

@@ -0,0 +1,52 @@
+#include <iostream>
+#include <omp.h>
+#include <random>
+
+using namespace std;
+
+int main() {
+  unsigned int seed = 0;
+  default_random_engine re{seed};
+  uniform_real_distribution<double> zero_to_one{0.0, 1.0};
+
+  int n = 100000000; // number of points to generate
+  int counter = 0; // counter for points lying in the first quadrant of a unit circle
+  auto start_time = omp_get_wtime(); // omp_get_wtime() is an OpenMP library routine
+
+#pragma omp parallel 
+  {
+    int num_threads = omp_get_num_threads();
+    int thread_id = omp_get_thread_num();
+    
+    int thread_shift = n / num_threads; // the remainder will just get handled by the last thread
+    int thread_start = thread_shift*thread_id;
+    int thread_end = thread_id == num_threads-1 ? n : thread_shift*(thread_id+1);
+#pragma omp critical
+    { // Just a pretty lil debug print out to make sure my division logic is logicing
+    std::cout << "Thread " << thread_id << " of " << num_threads << ": Computing " << thread_start << " to " << thread_end << std::endl;
+    }
+    
+    int thread_counter = 0;
+    
+    // compute n points and test if they lie within the first quadrant of a unit circle
+    for (int i = thread_start; i < thread_end; i++) {
+      auto x = zero_to_one(re); // generate random number between 0.0 and 1.0
+      auto y = zero_to_one(re); // generate random number between 0.0 and 1.0
+      if (x * x + y * y <= 1.0) { // if the point lies in the first quadrant of a unit circle
+        thread_counter++;
+      }
+    }
+    
+#pragma omp critical
+    {
+      counter = counter+thread_counter;
+    }
+  }
+
+  auto run_time = omp_get_wtime() - start_time;
+  auto pi = 4 * (double(counter) / n);
+
+  cout << "pi: " << pi << endl;
+  cout << "run_time: " << run_time << " s" << endl;
+  cout << "n: " << n << endl; 
+}