FEAT: Remap function can return nan and inf, Safe overload needed

[Shadowblitz16]

Tested versions

Reproducible in v4.4.stable.mono.official [e36cb51]

System information

Operating System: Debian GNU/Linux 13 KDE Plasma Version: 6.3.6 KDE Frameworks Version: 6.13.0 Qt Version: 6.8.2 Kernel Version: 6.12.57+deb13-amd64 (64-bit) Graphics Platform: Wayland Processors: 16 × AMD Ryzen 7 5700G with Radeon Graphics Memory: 16 GiB of RAM (11.6 GiB usable) Graphics Processor: AMD Radeon Graphics

Issue description

The remap math function returns nan and inf in valid cases making the function useless.

Steps to reproduce

Look at this code after you download the project and run it...

	var animation_speed := remap(player.velocity.x, 0.0, signf(player.direction) * walk_velocity_max, 0.0, 1.0)
	if  animation_speed == NAN: animation_speed = 0.0
	print(animation_speed)

Minimal reproduction project (MRP)

zelda-aol-arena.zip

[Arctis-Fireblight]

Thanks for the report, Can you let me know if this still happens in the 26.1 beta? https://github.com/Redot-Engine/redot-engine/releases/tag/redot-26.1-beta.0

[Shadowblitz16]

Thanks for the report, Can you let me know if this still happens in the 26.1 beta? https://github.com/Redot-Engine/redot-engine/releases/tag/redot-26.1-beta.0

I found the cause, it's because the function divides so if the right hand of the division is zero it fails.

I can test later.

[Arctis-Fireblight]

I found the cause, it's because the function divides so if the right hand of the division is zero it fails.

If the inputs cause a division by zero, wouldnt that indicate that inputs are invalid and should result in an error condition like NAN or INF?

[Shadowblitz16]

I found the cause, it's because the function divides so if the right hand of the division is zero it fails.

If the inputs cause a division by zero, wouldnt that indicate that inputs are invalid and should result in an error condition like NAN or INF?

Yes but it makes the function less useful because you have to check if imin and imax are the same beforehand.

What I would do is have a extra parameter that tells the function to return omin or omax when imin and imax are the same.

Something like MIN_IF_ZERO and MAX_IF_ZERO

Probably should be a enum.

[Arctis-Fireblight]

What I would do is have a extra parameter that tells the function to return omin or omax when imin and imax are the same.

Something like MIN_IF_ZERO and MAX_IF_ZERO

Probably should be a enum.

I am not sure about changing the default behavior of remap, but I'd be open to potentially make an overload that allows the user to specify if they want omin or omax on failure.

[Shadowblitz16]

What I would do is have a extra parameter that tells the function to return omin or omax when imin and imax are the same.
Something like MIN_IF_ZERO and MAX_IF_ZERO
Probably should be a enum.

I am not sure about changing the default behavior of remap, but I'd be open to potentially make an overload that allows the user to specify if they want omin or omax on failure.

As long as there are type safe overloads.

[Arctis-Fireblight]

So what I will do, is mark this down as a feature request for a safe overload for 'remap()` and tentitively mark this down for 26.2 inclusion

[Shadowblitz16]

So what I will do, is mark this down as a feature request for a safe overload for 'remap()` and tentitively mark this down for 26.2 inclusion

Ty