Refactor benchmark to use randomized vectors instead of sets

- Replaced std::set with std::vector for sequence handling.
- Added randomization of sequences to avoid ordered input bias.
- Removed unused balance function and related code in bst.cpp and bst.h.
- Fixed bugs in list insertion and search logic.
- Updated plot.py to allow custom y-axis labels and enable log scale for build plots.

Lab2/benchmark.cpp

@@ -2,6 +2,7 @@
 #include "bst/bst.h"
 #include "list/list.h"
 
+#include <algorithm>
 #include <chrono>
 #include <cstdio>
 #include <iostream>
@@ -9,7 +10,7 @@
 #include <random>
 #include <set>
 
-void measureList(std::set<int> *sequence, FILE *file) {
+void measureList(std::vector<int> *sequence, FILE *file) {
   float buildTime = 0, searchTime = 0, deleteTime = 0;
 
   for (int i = 0; i < 10; i++) {
@@ -55,7 +56,7 @@ void measureList(std::set<int> *sequence, FILE *file) {
   fprintf(file, "List,%f,%f,%f\n", buildTime, searchTime, deleteTime);
 }
 
-void measureBST(std::set<int> *sequence, FILE *file) {
+void measureBST(std::vector<int> *sequence, FILE *file) {
   float buildTime = 0, searchTime = 0, deleteTime = 0;
 
   for (int i = 0; i < 10; i++) {
@@ -68,7 +69,7 @@ void measureBST(std::set<int> *sequence, FILE *file) {
     for (int value : *sequence) {
       root = insert(root, value);
     }
-    root = balance(root);
+    // root = balance(root);
     auto end = std::chrono::high_resolution_clock::now();
     std::chrono::duration<double> elapsed_seconds = end - start;
     buildTime += elapsed_seconds.count() * 1000; // Convert to milliseconds
@@ -100,7 +101,7 @@ void measureBST(std::set<int> *sequence, FILE *file) {
   fprintf(file, "BST,%f,%f,%f\n", buildTime, searchTime, deleteTime);
 }
 
-void benchmarkAVL(std::set<int> *sequence, FILE *file) {
+void benchmarkAVL(std::vector<int> *sequence, FILE *file) {
   Tree *bst = nullptr;
   for (int value : *sequence) {
     bst = insert(bst, value);
@@ -147,19 +148,28 @@ int main() {
   }
 
   for (int n = 1; n < 26; n++) {
+    // Using a set here ensures that there are no duplicates
     std::set<int> sequence;
     while (sequence.size() < n * 1000) {
       sequence.insert(dis(gen));
     }
 
-    // Display this like a cascade
+    std::vector<int> random_sequence_vec;
+    for (int val : sequence) {
+        random_sequence_vec.push_back(val);
+    }
+
+    // Then randomize the sequence so the bst isn't a list (Set keeps the elements in order)
+    std::shuffle(random_sequence_vec.begin(), random_sequence_vec.end(), gen);
+
+    // Display the times like a cascade
     std::cout << "Running tests for " << n * 1000 << " elements..."
               << std::endl;
     if (mode == 1) {
-      measureList(&sequence, file);
-      measureBST(&sequence, file);
+      measureList(&random_sequence_vec, file);
+      measureBST(&random_sequence_vec, file);
     } else {
-      benchmarkAVL(&sequence, file);
+      benchmarkAVL(&random_sequence_vec, file);
     }
   }