What is range_lerp() ? What are it's function ?

ari484

Megalomaniak

from http://docs.godotengine.org/ko/latest/classes/class@gdscript.html#class-gdscript-range-lerp_ Maps a value from range [istart, istop] to [ostart, ostop]. range_lerp(75, 0, 100, -1, 1) # returns 0.5

And it was added in this commit: https://github.com/godotengine/godot/commit/d28da86f9ff5a70284e4a2078fa08867d4858a57

OK, so lets look at the above example again: range_lerp(75, 0, 100, -1, 1) # returns 0.5

The 75 is the integer value plugged into it, and then two ranges are given to it. If we look at the ranges given, 0 to 100 and -1 to 1 here are stated to be equivalent so the input of 75 is effectively then interpreted as 0.5(since that is three fourths, or 75 percent of the way linearly from -1 to 1) which is the value returned by the function.

As for the term lerp, it is short for linear interpolation.

Erich_L

Would it be too much to ask for someone to share what range_lerp would look like in mathematical notation?

Edit: range_lerp(value, min1, min2, max1, max2):

	min2 + (max2 - min2) * ((value - min1) / (max1 - min1))

xyz

@Erich_L said: Would it be too much to ask for someone to share what range_lerp would look like in mathematical notation?

Imho, much better name for range_lerp would be remap. Because it remaps the value from one min/max range to another.

Math is rather simple:

func range_lerp(value, min1, max1, min2, max2):
	var value_norm = inverse_lerp(min1, max1, value)
	return lerp( min2, max2, value_norm)

while lerp and inverse_lerp are:

func lerp(min, max, parameter):
	return min + (max-min) * parameter

func inverse_lerp( min, max, value):
	return (value-min) / (max-min)

Erich_L

Thankyou @xyz, I think by

    var value_norm = inverse_lerp(min1, min2, value)

you meant

    var value_norm = inverse_lerp(min1, max1, value)

xyz

@Erich_L Yep, good catch! Corrected it now.

Erich_L

@xyz It really isn't that complicated but it was a bit much for my brain to chew on, so thanks again this helped me a lot. Also, interestingly, even though I have min1 and max2 set to zero and therefore the equation can be simplified significantly, I found that continuing to use range_lerp() was five percent faster than the simplified version.

xyz

Sure. It's always better to let native code do the grunt work. Adding two numbers in native code is much faster than doing so in the script.

Erich_L

I stand corrected, I had a range lerp with: min1 and min2 as zero and max 2 as one. Simplification in the math saved me 139 milliseconds over 22,500 computations. :p

xyz

Always good to hear when leaps and bounds are made in the noble field of premature optimization :joy:

Erich_L

I've been range_lerping for lines and lines- for my terrain noise algorithm that I (you) have just about finished. It's about done so any tiny speed increases that I can put in my research results are... good? You're right remap would be a better name but I'm so used to range_lerp() now I'm over it. It's almost a little fun trying to convert code to math.

OpinionatedGamer

@Erich_L said: I've been range_lerping for lines and lines- for my terrain noise algorithm that I (you) have just about finished. It's about done so any tiny speed increases that I can put in my research results are... good? You're right remap would be a better name but I'm so used to range_lerp() now I'm over it. It's almost a little fun trying to convert code to math.

have you tried turning it to binary numbers yet?

Erich_L

@OpinionatedGamer I have not. Please tell me more. =)

xyz

Tbh, the term lerp irritates me :) And when I see inverse_lerp in code, my brain just freezes and I cannot comprehend what's going on. Not joking at all. I start scratching my head and reach for pen and paper to try to resolve it step by step. It gets me every time.

When working in C I tend to use my own functions. The names are: interpolate = lerp normalize = inverse_lerp remap = range_lerp With these names I know exactly what is going on.

~ Lerping is one thing, but don't mess with binary at home on your own, fellow kids. It's dangerous. People you've seen doing it online were trained professionals.