Spatial Vectors

Minetest stores 3-dimensional spatial vectors in Lua as tables of 3 coordinates, and has a class to represent them (vector.*), which this chapter is about. For details on what a spatial vectors is, please refer to Wikipedia: https://en.wikipedia.org/wiki/Euclidean_vector.

Spatial vectors are used for various things, including, but not limited to:

  • any 3D spatial vector (x/y/z-directions)
  • Euler angles (pitch/yaw/roll in radians) (Spatial vectors have no real semantic meaning here. Therefore, most vector operations make no sense in this use case.)

Note that they are not used for:

  • n-dimensional vectors where n is not 3 (ie. n=2)
  • arrays of the form {num, num, num}

The API documentation may refer to spatial vectors, as produced by vector.new, by any of the following notations:

  • (x, y, z) (Used rarely, and only if it's clear that it's a vector.)
  • vector.new(x, y, z)
  • {x=num, y=num, z=num} (Even here you are still supposed to use vector.new.)

Compatibility notes

Vectors used to be defined as tables of the form {x = num, y = num, z = num}. Since Minetest 5.5.0, vectors additionally have a metatable to enable easier use. Note: Those old-style vectors can still be found in old mod code. Hence, mod and engine APIs still need to be able to cope with them in many places.

Manually constructed tables are deprecated and highly discouraged. This interface should be used to ensure seamless compatibility between mods and the Minetest API. This is especially important to callback function parameters and functions overwritten by mods. Also, though not likely, the internal implementation of a vector might change in the future. In your own code, or if you define your own API, you can, of course, still use other representations of vectors.

Vectors provided by API functions will provide an instance of this class if not stated otherwise. Mods should adapt this for convenience reasons.

Special properties of the class

Vectors can be indexed with numbers and allow method and operator syntax.

All these forms of addressing a vector v are valid: v[1], v[3], v.x, v[1] = 42, v.y = 13 Note: Prefer letter over number indexing for performance and compatibility reasons.

Where v is a vector and foo stands for any function name, v:foo(...) does the same as vector.foo(v, ...), apart from deprecated functionality.

tostring is defined for vectors, see vector.to_string.

The metatable that is used for vectors can be accessed via vector.metatable. Do not modify it!

All vector.* functions allow vectors {x = X, y = Y, z = Z} without metatables. Returned vectors always have a metatable set.

Common functions and methods

For the following functions (and subchapters), v, v1, v2 are vectors, p1, p2 are position vectors, s is a scalar (a number), vectors are written like this: (x, y, z):

  • vector.new([a[, b, c]]):
    • Returns a new vector (a, b, c).
    • Deprecated: vector.new() does the same as vector.zero() and vector.new(v) does the same as vector.copy(v)
  • vector.zero():
    • Returns a new vector (0, 0, 0).
  • vector.copy(v):
    • Returns a copy of the vector v.
  • vector.from_string(s[, init]):
    • Returns v, np, where v is a vector read from the given string s and np is the next position in the string after the vector.
    • Returns nil on failure.
    • s: Has to begin with a substring of the form "(x, y, z)". Additional spaces, leaving away commas and adding an additional comma to the end is allowed.
    • init: If given starts looking for the vector at this string index.
  • vector.to_string(v):
    • Returns a string of the form "(x, y, z)".
    • tostring(v) does the same.
  • vector.direction(p1, p2):
    • Returns a vector of length 1 with direction p1 to p2.
    • If p1 and p2 are identical, returns (0, 0, 0).
  • vector.distance(p1, p2):
    • Returns zero or a positive number, the distance between p1 and p2.
  • vector.length(v):
    • Returns zero or a positive number, the length of vector v.
  • vector.normalize(v):
    • Returns a vector of length 1 with direction of vector v.
    • If v has zero length, returns (0, 0, 0).
  • vector.floor(v):
    • Returns a vector, each dimension rounded down.
  • vector.round(v):
    • Returns a vector, each dimension rounded to nearest integer.
    • At a multiple of 0.5, rounds away from zero.
  • vector.apply(v, func):
    • Returns a vector where the function func has been applied to each component.
  • vector.combine(v, w, func):
    • Returns a vector where the function func has combined both components of v and w for each component
  • vector.equals(v1, v2):
    • Returns a boolean, true if the vectors are identical.
  • vector.sort(v1, v2):
    • Returns in order minp, maxp vectors of the cuboid defined by v1, v2.
  • vector.angle(v1, v2):
    • Returns the angle between v1 and v2 in radians.
  • vector.dot(v1, v2):
    • Returns the dot product of v1 and v2.
  • vector.cross(v1, v2):
    • Returns the cross product of v1 and v2.
  • vector.offset(v, x, y, z):
    • Returns the sum of the vectors v and (x, y, z).
  • vector.check(v):
    • Returns a boolean value indicating whether v is a real vector, eg. created by a vector.* function.
    • Returns false for anything else, including tables like {x=3,y=1,z=4}.
  • vector.in_area(pos, min, max):
    • Returns a boolean value indicating if pos is inside area formed by min and max.
    • min and max are inclusive.
    • If min is bigger than max on some axis, function always returns false.
    • You can use vector.sort if you have two vectors and don't know which are the minimum and the maximum.

For the following functions x can be either a vector or a number:

  • vector.add(v, x):
    • Returns a vector.
    • If x is a vector: Returns the sum of v and x.
    • If x is a number: Adds x to each component of v.
  • vector.subtract(v, x):
    • Returns a vector.
    • If x is a vector: Returns the difference of v subtracted by x.
    • If x is a number: Subtracts x from each component of v.
  • vector.multiply(v, s):
    • Returns a scaled vector.
    • Deprecated: If s is a vector: Returns the Schur product.
  • vector.divide(v, s):
    • Returns a scaled vector.
    • Deprecated: If s is a vector: Returns the Schur quotient.

Operators

Operators can be used if all of the involved vectors have metatables: * v1 == v2: * Returns whether v1 and v2 are identical. * -v: * Returns the additive inverse of v. * v1 + v2: * Returns the sum of both vectors. * Note: + cannot be used together with scalars. * v1 - v2: * Returns the difference of v1 subtracted by v2. * Note: - cannot be used together with scalars. * v * s or s * v: * Returns v scaled by s. * v / s: * Returns v scaled by 1 / s.

For the following functions a is an angle in radians and r is a rotation vector ({x = <pitch>, y = <yaw>, z = <roll>}) where pitch, yaw and roll are angles in radians.

  • vector.rotate(v, r):
    • Applies the rotation r to v and returns the result.
    • vector.rotate(vector.new(0, 0, 1), r) and vector.rotate(vector.new(0, 1, 0), r) return vectors pointing forward and up relative to an entity's rotation r.
  • vector.rotate_around_axis(v1, v2, a):
    • Returns v1 rotated around axis v2 by a radians according to the right hand rule.
  • vector.dir_to_rotation(direction[, up]):
    • Returns a rotation vector for direction pointing forward using up as the up vector.
    • If up is omitted, the roll of the returned vector defaults to zero.
    • Otherwise direction and up need to be vectors in a 90 degree angle to each other.

Further helpers

There are more helper functions involving vectors, but they are listed elsewhere because they only work on specific sorts of vectors or involve things that are not vectors.

For example:

  • minetest.hash_node_position (Only works on node positions.)
  • minetest.dir_to_wallmounted (Involves wallmounted param2 values.)