SQUINT (Static Quantities in Tensors)

SQUINT is a header-only C++ library designed for compile-time dimensional analysis, unit conversion, and linear algebra operations. It's particularly well-suited for graphics programming and physics simulations, combining a quantity system for handling physical units and dimensions with a tensor system for efficient numerical computations.

Key Features

• Compile-time dimensional analysis
• Flexible tensor system supporting both fixed and dynamic shapes
• Integration of physical quantities with tensor operations
• Optional runtime error checking
• Support for common linear algebra operations
• Useful mathematical and physical constants

Installation

SQUINT is a header-only library. To use it in your project:

1. Copy the include/squint directory to your project's include path.
2. Include the necessary headers in your C++ files:

#include <squint/quantity.hpp>
#include <squint/tensor.hpp>
#include <squint/geometry.hpp>

For CMake projects, you can use FetchContent for a more streamlined integration:

include(FetchContent)

FetchContent_Declare(
    squint
    GIT_REPOSITORY https://github.com/barne856/squint.git
    GIT_TAG main  # or a specific tag/commit
)

FetchContent_MakeAvailable(squint)

target_link_libraries(your_target PRIVATE SQUINT::SQUINT)

Examples

SQUINT can be used for common graphics operations:

#include <iostream>
#include <squint/geometry.hpp>
#include <squint/quantity.hpp>
#include <squint/tensor.hpp>

using namespace squint;
using namespace squint::units::literals;

int main() {
    // Define a 3D point
    vec3_t<length> point{
      1.0_m,
      2.0_m,
      3.0_m
    };

    // Create a model matrix
    mat4 model = mat4::eye();

    // Apply transformations
    geometry::translate(
    model,
    vec3_t<length>{
        2.0_m,
        1.0_m,
        0.0_m
    });
    geometry::rotate(
    model,
    math_constants<float>::pi / 4.0f,
    vec3{
        0.0f,
        1.0f,
        0.0f
    });
    geometry::scale(
    model, 
    vec3{
        2.0f,
        2.0f,
        2.0f
    });

    // Transform the point
    vec4_t<length> homogeneous_point{point(0), point(1), point(2), length(1.0f)};
    auto transformed_point = model * homogeneous_point;

    std::cout << "point: " << point << std::endl;
    std::cout << "transformed_point: " << transformed_point << std::endl;

  return 0;
}

SQUINT can be used for physics calculations:

#include <iostream>
#include <squint/quantity.hpp>
#include <squint/tensor.hpp>

using namespace squint;

int main() {
  // Define initial conditions
  auto pos = vec3_t<length>::zeros();
  vec3_t<velocity> vel{velocity(5.0f), velocity(10.0f), velocity(0.0f)};
  vec3_t<acceleration_t<float>> acc{acceleration(0.0f), acceleration(-9.81f),
                                    acceleration(0.0f)};

  // Simulation parameters
  auto dt = squint::units::seconds(0.1f);
  auto total_time = squint::units::seconds(1.0f);

  // Simulation loop
  for (auto t = squint::units::seconds(0.0f); t < total_time; t += dt) {
    // Update position and velocity
    pos += vel * dt + 0.5f * acc * dt * dt;
    vel += acc * dt;

    // Print current state
    std::cout << "Time: " << t << "\nPosition: " << pos
              << "\nVelocity: " << vel << "\n\n";
  }

  return 0;
}

Documentation

For detailed information about SQUINT's features, API reference, and advanced usage, please refer to the full documentation.

License

MIT