Scientific Computing

By: Tadhg Jones, David Kraakman, Storm Koper
Team: 5

Repository Structure

├── pyproject.toml           # Project configuration and dependencies
├── README.md                # Project documentation and commands
└── set_1/                   # Code for Assignment Set 1
    ├── scripts/             # Executable scripts to generate plots and run experiments
    │   ├── benchmarks.py      # Benchmarking tool for base vs JIT performance
    │   ├── run_tide.py        # Scripts for Time-Independent Diffusion (Questions H-L)
    │   ├── run_wave1D.py      # Scripts for 1D Wave Equation (Questions B-C)
    │   └── run_wave2D.py      # Scripts for Time dependent 2D Diffusion Equation (Questions E-G)
    ├── tests/               # Pytest unit tests
    │   ├── test_misc.py       # Tests for helper functions
    │   ├── test_tide.py       # Tests for TIDE solvers (Jacobi, Gauss-Seidel, SOR)
    │   └── test_wave.py       # Tests for Wave solvers (1D, 2D, Leapfrog)
    └── utils/               # Core numerical solvers and helper modules
        ├── config.py          # Configuring global styles
        ├── misc.py            # Analytical solutions and image loading utilities
        ├── optimized.py       # Numba JIT-compiled optimized numerical schemes
        ├── TIDE.py            # Time-Independent Diffusion Equation solvers
        └── wave.py            # Wave Equation solvers

# Code for Assignment Set 2

This folder contains the code for Assignment Set 2. 

## Structure

```text
├── figures/                 # Output figures and animations
├── scripts/                 # Executable scripts to generate plots and run experiments
│   ├── main.py                # Main entrypoint to generate all report figures and animations
│   ├── run_dla.py             # Scripts for Diffusion-Limited Aggregation (DLA)
│   ├── run_mc.py              # Scripts for Monte Carlo DLA
│   └── run_rd.py              # Scripts for Reaction-Diffusion (Gray-Scott model)
├── tests/                   # Pytest unit tests
│   ├── test_dla.py            # Tests for DLA solvers
│   ├── test_mc.py             # Tests for Monte Carlo implementations
│   └── test_rd.py             # Tests for Reaction-Diffusion solvers
└── utils/                   # Core numerical solvers and helper modules
    ├── config.py              # Configuring global styles
    ├── DLA.py                 # Diffusion-Limited Aggregation solvers
    ├── MC_DLA.py              # Monte Carlo routines
    └── RD.py                  # Reaction-Diffusion solvers

# Code for Assignment Set 3

This folder contains the code for Assignment Set 3.

## Structure

```text
├── figures/                 # Output figures
├── scripts/                 # Executable scripts to generate plots and run experiments
│   ├── benchmark_fd.py        # Benchmarking tool for finite difference
│   ├── run_fd.py              # Scripts running Finite Difference
│   ├── run_FE.py              # Scripts running Finite Element
│   ├── run_LB.py              # Scripts running Lattice Boltzman
│   └── run_LBA.py             # Outdated
└── utils/                   # Core numerical solvers and helper modules
    ├── config.py              # Configuring global styles
    ├── FD.py                  # Finite Difference
    ├── FE.py                  # Finite Element
    ├── LB.py                  # Lattice Boltzmann
    └── WIFI.py                # WiFi-Optimization

Running the Code

The main entrypoint for set 2 is scripts/main.py. Running this file can be executed in two ways:

1. Generate Report Figures

   uv run -m set_2.scripts.main

   This will create all the figures used in the report. These will be outputted in a popup window as well as saved in figures/. During the execution, the terminal will display what it is working on currently.

2. Generate Animations

   uv run -m set_2.scripts.main --animations

   This will NOT create any of the figures used in the report, but instead output some animations we made for the DLA and Gray-Scott model.

(Note: If you are not using uv, you can replace uv run with python after activating your virtual environment).

Tests

We have also included some tests for this project, which can be run through:

uv run pytest set_2

Dependency Management

For this project we use uv for package and virtual environment management. If you have uv installed, you can set up the environment and install dependencies by running:

uv sync

However, if you use another package and virtual environt manager, then you can use the provided pyproject.toml. For example, using standard pip:

python -m venv .venv
source .venv/bin/activate  # On Windows use: .venv\Scripts\activate
pip install .

Set 1

Below are the CLI commands to obtain the figures used in the report for set 1. Note that if you are not using uv you can replace uv run with python after activating your virtual environment.

1D Wave Equation

• Question B (Time development of 1D wave):

  uv run -m set_1.scripts.run_wave1D -b_type i -steps 1000 -save_every 10
  uv run -m set_1.scripts.run_wave1D -b_type ii -steps 1000 -save_every 10
  uv run -m set_1.scripts.run_wave1D -b_type iii -steps 1000 -save_every 10

• Question C (Animated Wave):

  uv run -m set_1.scripts.run_wave1D -b_type i -steps 1000 -save_every 10 --animate
  uv run -m set_1.scripts.run_wave1D -b_type ii -steps 1000 -save_every 10 --animate
  uv run -m set_1.scripts.run_wave1D -b_type iii -steps 1000 -save_every 10 --animate

• Optional question (Leapfrog):

  uv run -m set_1.scripts.run_wave1D -b_type i -steps 5000 --compare
  uv run -m set_1.scripts.run_wave1D -b_type iii -steps 5000 --compare

2D Diffusion

• Question E (2D Diffusion vs Analytical Solution)

  uv run -m set_1.scripts.run_wave2D -question E

• Question F (2D domain at various time steps)

  uv run -m set_1.scripts.run_wave2D -question F

• Question G (Animation of 2D domain)

  uv run -m set_1.scripts.run_wave2D -question G

Time-Independent Diffusion Equation (TIDE)

• Question H (3 iteration methods vs analytical solution)

  uv run -m set_1.scripts.run_tide -question H

• Question I (2D domain at various time steps)

  uv run -m set_1.scripts.run_tide -question I

• Question J (Omega sweep + golden section search for empty grid)

  uv run -m set_1.scripts.run_tide -question J

• Question K (Time evolution of diffusion with sinks + optimal omega)

  uv run -m set_1.scripts.run_tide -question K

• Question L (Time evolution of diffusion with insulators + optimal omega)

  uv run -m set_1.scripts.run_tide -question L

Misc

• Run tests with:

  uv run pytest set_1

• Run benchmarks with:

  uv run -m set_1.scripts.benchmarks -method SOR -iterations 1000 -N 500 -repeats 5 --warmup_jit

Set 2

The main entrypoint for set 2 is main.py. Running this file can be executed in two ways (Note that if you are not using uv you can replace uv run with python after activating your virtual environment).

1. Generate Report Figures

   uv run -m set_2.scripts.main

   This will create all the figures used in the report. These will be outputted in a popup window as well as saved in figures/. During the execution, the terminal will display what it is working on currently.

2. Generate Animations

   uv run -m set_2.scripts.main --animations

   This will NOT create any of the figures used in the report, but instead output some nice (non-required) animations we made for the DLA and Gray-Scott model.

Tests

We have also included some tests for this project, which can be ran through:

uv run pytest set_2

Set 3

The main entrypoint for set 2 is main.py. Running this file can be executed in two ways (Note that if you are not using uv you can replace uv run with python after activating your virtual environment).

Challenge A: Fluid Dynamics

• Generating the Divergence and Strouhal plots:

  uv run -m set_3.scripts.run_fd
  uv run -m set_3.scripts.run_LB
  uv run -m set_3.scripts.run_FE

Challenge A: WiFi-Optimization

• Performing the raw signal optimisation (note, the full analysis takes ~40 minutes, choose --analysis 1, to just run shorter version with only 1 optimalisation run.):

  uv run -m set_3.utils.WIFI --analysis 5 --raw

• Performing the dB signal optimisation (note, the full analysis takes ~40 minutes, choose --analysis 1, to just run shorter version with only 1 optimalisation run.):

  uv run -m set_3.utils.WIFI --analysis 5