Library for extremely lightweight vector and matrix operations.
To use this package, run the following command in your project's root directory:
Put the following dependency into your project's dependences section:
dvec - Easy linear algebra for D
dvec is a library that offers vector and matrix data structures and functions in a way that's intuitive, easy to use, tested, and well-documented. It takes advantage of the D language's advanced templating system and compile-time code-generation to help optimize performance and type safety.
dvec offers the basic linear algebra primitives:
Each of these can be of any desired integer size (within the limits of your system). For example,
Vec!(float, 5) v; defines a 5-dimensional floating-point vector, and
Mat!(int, 2, 3) m; defines a 2x3 (2 rows by 3 columns) integer matrix. However, this can get to be a bit tedious, so we've included the most common vectors and matrices as aliases for you:
- 2, 3, and 4-dimensional vectors of float, double, and int types:
Vec4f, for example.
- 2, 3, and 4-dimensional square matrices of float, double, and int types:
Mat4d, for example.
Each primitive is a simple struct with a fully transparent internal structure, and implements all the basic arithmetic operators, so you can do things like this easily:
Vec3f v = Vec3f(1f, 2f, 3f) * 5 - Vec3f(0.5f); // v.data == [4.5f, 9.5f, 14.5f]
Besides the `toString()` methods, no parts of this library make use of the GC, and are thus
Vectors are simply a fixed-size list of numerical elements, which implement a lot of mathematical operations that are useful in linear algebra.
The values of a vector can be accessed in the following ways:
- By index:
vgets the second element in vector
- By name:
v.x, v.y, v.z, v.wgets the first, second, third, or fourth element of a vector, respectively.
- By the internal data:
v.datagets the internal array.
You can cast vectors of different types and sizes using D's standard
cast(type) syntax. Note that the individual elements are casted using
cast as well, and that may result in data loss if you, for example, cast from a floating-point to integer vector. If you cast to a vector with a different size, the following rules apply:
- If the new vector type is smaller, extra elements are truncated.
- If the new vector type is larger, any missing elements are initialized to zero.
Certain vector types get access to extra special functions:
- Any floating-point (float, double) vector supports the
norm()method, to normalize the vector to a unit vector of length 1.
- Any floating-point (float, double) 2-dimensional vector supports the
toCartesian()methods, so that the vector can be treated as polar or cartesian coordinates and converted between the two.
- Any floating-point (float, double) 3-dimensional vector supports the
cross()method for computing the cross product with another 3-dimensional vector.
Certain matrix types get access to extra special functions:
- Square matrices (NxM matrices where N = M) support a variety of additional methods that are useful for linear algebra, like
inv()(inverse), and so on.
- 3x3 matrices support linear transformation methods in 2 dimensions, like translation, rotation, skew, and scaling.
- 4x4 matrices support linear transformation methods in 3 dimensions.
For more information, please see dub project page and its associated documentation.
A lot of effort has gone into making sure everything in this library is well-documented and explained well. If you find that documentation is missing or insufficient for a certain part, please make an issue on GitHub!
- Registered by Andrew Lalis
- 0.0.7 released 11 months ago
- Copyright © 2022, Andrew Lalis