diff --git a/README.md b/README.md
index 28faabf..5754785 100644
--- a/README.md
+++ b/README.md
@@ -17,6 +17,12 @@ python utils/correctness.py --alg brute --csv tiny
 module load python
 python utils/visualizer.py tiny build/brute.out
 
+# For Timing Comparison Graph
+
+module load python
+python utils/comparisons.py (tries all implementations)
+python utils/comparisons.py greedy dp (only compares greedy and dp)
+
 # References
 
 https://www.kaggle.com/datasets/mexwell/traveling-salesman-problem/data
\ No newline at end of file
diff --git a/algorithms/greedy.cpp b/algorithms/greedy.cpp
index e69de29..8a9154f 100644
--- a/algorithms/greedy.cpp
+++ b/algorithms/greedy.cpp
@@ -0,0 +1,59 @@
+#include <bits/stdc++.h>
+#include "../common/algorithms.hpp"
+
+using namespace std;
+
+// Helper function to calculate Euclidean distance
+double distance(const std::pair<double, double>& p1, const std::pair<double, double>& p2) {
+    return std::sqrt(std::pow(p1.first - p2.first, 2) + std::pow(p1.second - p2.second, 2));
+}
+
+TSPResult solve(const std::vector<std::pair<double, double>>& coordinates) {
+    int n = coordinates.size();
+    
+    // Compute distance matrix
+    vector<vector<double>> distances(n, vector<double>(n));
+    for (int i = 0; i < n; i++) {
+        for (int j = 0; j < n; j++) {
+            distances[i][j] = distance(coordinates[i], coordinates[j]);
+        }
+    }
+
+    // Initialize variables for greedy algorithm
+    vector<bool> visited(n, false);
+    vector<int> path;
+    double total_cost = 0.0;
+    
+    // Start from city 0
+    int current_city = 0;
+    path.push_back(current_city);
+    visited[current_city] = true;
+
+    // Visit remaining n-1 cities
+    while (path.size() < n) {
+        double min_distance = numeric_limits<double>::infinity();
+        int next_city = -1;
+
+        // Find nearest unvisited city
+        for (int j = 0; j < n; j++) {
+            if (!visited[j] && distances[current_city][j] < min_distance) {
+                min_distance = distances[current_city][j];
+                next_city = j;
+            }
+        }
+
+        // Add nearest city to path
+        current_city = next_city;
+        path.push_back(current_city);
+        visited[current_city] = true;
+        total_cost += min_distance;
+    }
+
+    // Add cost of returning to start
+    total_cost += distances[path.back()][path[0]];
+
+    TSPResult result;
+    result.cost = total_cost;
+    result.path = path;
+    return result;
+}
\ No newline at end of file
diff --git a/utils/comparisons.py b/utils/comparisons.py
new file mode 100644
index 0000000..3032e20
--- /dev/null
+++ b/utils/comparisons.py
@@ -0,0 +1,88 @@
+#!/usr/bin/env python3
+import subprocess
+import time
+import argparse
+import matplotlib.pyplot as plt
+
+def run_implementation(exec_path: str, dataset: str) -> float:
+    """Run implementation with dataset and return execution time"""
+    try:
+        start = time.time()
+        result = subprocess.run([exec_path, '--csv', f'data/{dataset}.csv'], 
+                              stdout=subprocess.PIPE,
+                              stderr=subprocess.PIPE)
+        end = time.time()
+        return end - start
+    except Exception as e:
+        print(f"Error running {exec_path} with {dataset}: {e}")
+        return -1
+
+def plot_results(results):
+    """Plot timing results for all implementations"""
+    plt.figure(figsize=(12, 8))
+    
+    colors = {
+        'brute': 'b',
+        'greedy': 'g',
+        'genetic': 'r',
+        'dp': 'purple',
+        'greedy_omp': 'y'
+    }
+    
+    datasets = ['tiny', 'small', 'medium', 'large','huge','gigantic']
+    x_pos = range(len(datasets))
+    
+    for impl, timings in results.items():
+        if timings:  # Only plot if we have data
+            plt.plot(x_pos, [timings[d] for d in datasets], 
+                    f'{colors[impl]}-o', 
+                    label=f'{impl.capitalize()} Implementation')
+    
+    plt.xticks(x_pos, datasets)
+    plt.xlabel('Dataset Size')
+    plt.ylabel('Execution Time (seconds)')
+    plt.title('TSP Implementation Performance Comparison')
+    plt.legend()
+    plt.grid(True)
+    plt.yscale('log')  # Log scale for better visibility
+    
+    plt.savefig('comparison_results.png')
+    print("Plot saved as comparison_results.png")
+    
+    try:
+        plt.show()
+    except Exception as e:
+        print(f"Could not display plot: {e}")
+
+if __name__ == "__main__":
+    all_implementations = ['brute', 'greedy', 'genetic', 'dp', 'greedy_omp']
+    
+    parser = argparse.ArgumentParser(description='Compare TSP implementations')
+    parser.add_argument('implementations', 
+                       nargs='*',
+                       choices=all_implementations,
+                       help='Implementations to compare. If none specified, runs all.')
+    
+    args = parser.parse_args()
+    
+    # Select implementations to run
+    implementations = args.implementations if args.implementations else all_implementations
+    datasets = ['tiny', 'small', 'medium', 'large','huge','gigantic']
+    
+    results = {}
+    
+    # Run tests for each implementation and dataset
+    for impl in implementations:
+        results[impl] = {}
+        exec_path = f'./build/{impl}'
+        
+        for dataset in datasets:
+            print(f"Running {impl} on {dataset} dataset...")
+            execution_time = run_implementation(exec_path, dataset)
+            
+            if execution_time >= 0:
+                results[impl][dataset] = execution_time
+                print(f"{impl.capitalize()}, {dataset}: {execution_time:.6f} seconds")
+    
+    # Plot results
+    plot_results(results)
\ No newline at end of file