CPP AP Project

---

Overview

This repository contains a C++ project structured in two main parts, each focusing on different aspects of custom data structures and object-oriented design. The first part explores the implementation and performance comparison of custom vector classes against the standard C++ std::vector. The second part involves the creation of a complex Device class, which models real-world objects by aggregating various shapes and sub-devices, each with their own geometric and physical properties.

The project emphasizes efficient memory management, performance benchmarking, and the practical application of object-oriented principles in C++.

---

Part 1: Vector Class Comparison

In this part, two custom vector implementations are compared:

1. Single Pointer Vector (T*): This implementation uses a single pointer for managing the underlying array.
2. Double Pointer Vector (T**): This implementation uses a double pointer for managing the underlying array.

Both custom implementations are benchmarked against std::vector from the C++ Standard Library.

Benchmark Details

The benchmarks were conducted using objects of varying sizes, from 100 * 1MB objects to 2 * 50MB objects, keeping the total vector size at 100MB.

• Results Overview: The std::vector consistently outperformed both custom implementations. The double pointer vector (T**) was close in performance to std::vector, while the single pointer vector (T*) lagged significantly behind.

• Performance Analysis:
  • The single pointer vector (T*) performance is more sensitive to the number of elements.
  • The double pointer vector (T**) performance is more sensitive to the size of individual objects.

• Detailed Comparison with std::vector:
  • In some cases, the double pointer vector (T**) even outperformed std::vector, but overall, std::vector was the faster choice. This might be due to std::vector’s capacity management, where additional memory is pre-allocated. However, when the allocated capacity is filled, std::vector can become slower.

---

Part 2: Device Class

In the second part, a Device class is implemented. This class models a realistic device, consisting of several shape objects as well as other sub-devices. The Device class and its components are designed to represent the complex structure of real-world objects, like an automobile with various parts, both simple and complex.

Functionalities

The Device class includes the following key functionalities:

• AddShape: Adds a shape to the device.
• AddSubDevice: Adds a sub-device to the device.
• GetVolume: Returns the total volume of the device, including all shapes and sub-devices. The volume is calculated by invoking the EvalVolume function whenever any shape's dimensions are modified using the Set functions.
• GetSurfaceArea: Returns the total surface area of the device, including all shapes and sub-devices. The surface area is calculated by invoking the EvalSurfaceArea function whenever any shape's dimensions are modified.
• GetMass: Returns the total mass of the device, including all shapes and sub-devices. The mass is calculated by invoking the EvalMass function whenever any shape's density or volume is modified.

Shape Classes

The Shape class and its derived classes (Cube, Sphere, Cylinder, Pyramid) include methods to calculate volume, surface area, and mass based on the shape's dimensions and density. These calculations are done via Eval functions, which are triggered whenever relevant properties are set or changed.

Additionally, a custom operator<< function is implemented to provide detailed output about the device and its components.

You can experiment with this implementation by running the sample code provided in the Part2 folder.

---

Compilation

The code can be compiled using g++ or MSVC.

For convenience, precompiled binaries are available:

• Download Part1.exe
• Download Part2.exe