vector.abs

[Oil_boi]

Here is a simple code that can be added into the lua code to return an absolute vector

Code: Select all

--this converts a vector into absolute value
function vector.abs(vector)
	vector.x = math.abs(vector.x)
	vector.y = math.abs(vector.y)
	vector.z = math.abs(vector.z)
	return(vector)
end

I can't find where the vector library is stored in the game engine so I'll just post this here and hope someone makes a github pull request to add it in ¯\_(ツ)_/¯

[sfan5]

I can't find where the vector library is stored in the game engine

Right here: https://github.com/minetest/minetest/bl ... vector.lua

[Hume2]

I can't find where the vector library is stored in the game engine

Right here: https://github.com/minetest/minetest/bl ... vector.lua

I think, you confused that with the function vector.length. What OP says is not any kind of norm. It only makes all coordinates positive.

[Oil_boi]

I can't find where the vector library is stored in the game engine

Right here: https://github.com/minetest/minetest/bl ... vector.lua

I think, you confused that with the function vector.length. What OP says is not any kind of norm. It only makes all coordinates positive.

This can be used for collision detection

Code: Select all

pos = vector.new(0,0,0)
pos2 = vector.new(0.5,0,0)

instead of checking like:

Code: Select all

diff = vector.subtract(pos,pos2)
if diff.x < 0.5 and diff.x > -0.5 and diff.z < 0.5 and diff.z > -0.5 do

you can check like

Code: Select all

diff = vector.abs(vector.subtract(pos,pos2))
if diff.x < 0.5 and diff.z < 0.5 then

I specifically utilize this function in these locations:
https://github.com/oilboi/Crafter/blob/ ... on.lua#L56
https://github.com/oilboi/Crafter/blob/ ... te.lua#L61
https://github.com/oilboi/Crafter/blob/ ... te.lua#L74

Also I did a pull request
https://github.com/minetest/minetest/pull/9551

[Hume2]

This can be actually done even easier by using the maximum norm:

Code: Select all

function vector.maxnorm(v)
    return math.max(math.abs(v.x), math.max(math.abs(v.y), math.abs(v.z)))
end

Then it reduces to:

Code: Select all

norm = vector.maxnorm(vector.multiply(vector.subtract(pos, pos2), {x=2, y=0, z=2}))
if norm < 1 then

[Oil_boi]

This can be actually done even easier by using the maximum norm:

Code: Select all

function vector.maxnorm(v)
    return math.max(math.abs(v.x), math.max(math.abs(v.y), math.abs(v.z)))
end

Then it reduces to:

Code: Select all

norm = vector.maxnorm(vector.multiply(vector.subtract(pos, pos2), {x=2, y=0, z=2}))
if norm < 1 then

Yes that is very good, but it's not modular as you can't recover the actual value of Y as it was before maxnorming it, but that would be a nice function to have

[AspireMint]

what about vector.distance(p1, p2) < 0.5

[Hume2]

what about vector.distance(p1, p2) < 0.5

This calculates the Eucleidan distance, not the maximum norm distance. Please, read the OP toughly to see what the vector.abs actually does.

[Oil_boi]

I just decided to make a video on it: https://youtu.be/2GOsXSeNgQw

[GreenXenith]

vector.apply(v, math.abs) but ok.

[Oil_boi]

vector.apply(v, math.abs) but ok.

Well I see a lot of people saying this is redundant with the vector.apply function, but, as they say that, we also have:

Code: Select all

function vector.floor(v)
	return {
		x = math.floor(v.x),
		y = math.floor(v.y),
		z = math.floor(v.z)
	}
end

function vector.round(v)
	return {
		x = math.floor(v.x + 0.5),
		y = math.floor(v.y + 0.5),
		z = math.floor(v.z + 0.5)
	}
end
function vector.add(a, b)
	if type(b) == "table" then
		return {x = a.x + b.x,
			y = a.y + b.y,
			z = a.z + b.z}
	else
		return {x = a.x + b,
			y = a.y + b,
			z = a.z + b}
	end
end

Why do we have these? Because it's easier to type

Code: Select all

vector.round(v) 

rather than

Code: Select all

vector.apply(v, function(x) return(math.floor(x + 0.5)) end)

and also

Code: Select all

vector.add(a, b)

instead of

Code: Select all

local new_vector = {x=a.x+b.x,y=a.y+b.y,z=a.z+b.z}

It makes more logical sense for the library to contain functions to apply a basic bit of math to each vector coordinate rather than each programmer to define their own functions.

We could even boil it down even further using the logic that is applied in the responses to this. Why even have vector.new when we could run:

Code: Select all

v = {x=5,y=1,z=4}

Very simply, because it's much easier to type out:

Code: Select all

v = vector.new(5,1,4)

And along with the logic of what I've been trying to provide here, it is also easier to type and remember:

Code: Select all

vector.abs(v)

I dunno how to explain it any further, it's only 6 lines of codes. It doesn't cause any performance changes to the engine. It makes your life slightly easier. If someone doesn't want it, they could keep doing vector.apply(v, math.abs). I also thought everyone would like the small addition.

I dunno man ¯\_(ツ)_/¯