Converting number ranges - GameCreators Forum

Author

Message

Redmotion

22

Years of Service

User Offline

Joined: 16th Jan 2003

Location: Mmm mmm.. Marmite

Posted: 22nd Sep 2007 13:02

Link

I want to convert a number that exists within the range:

0 to 65535

to exist within:

-100 to +100 (so the 32767 would become 0)

Is there a standard maths technique to do this and if so, what's it called?!

Something like:

out = convertrange(in,oldlow,oldhigh,newlow,newhigh)

Cheers in advance.

Check out The Grand Machines! & Black Mesa: Source

Back to top

Profile PM

revenant chaos

DBPro Master

18

Years of Service

User Offline

Joined: 21st Mar 2007

Location: Robbinsdale, MN

Posted: 22nd Sep 2007 13:27 Edited at: 22nd Sep 2007 13:40

Link

take your number, subtract 32767 from it, and multiply it by 0.0000305

example:

+ Code Snippet

currentvalue=32767
converted as float

do
cls
set cursor 0,0
print "current value=",currentvalue
converted=(currentvalue-32767)*0.0000305
print "the converted value=",converted

currentvalue=currentvalue+((upkey()-downkey())*10)
if currentvalue<0 then currentvalue=0
if currentvalue>65535 then currentvalue=65535

loop

press the up and down arrows to change the value.

to figure out the value (0.0000305) is really quite simple. so your range is 0 to 65535 and you want the lower half of it to be a negative. divide the maximum value by 2 (=32767). now you want this number to represent the number 1. divide the number 1 by 32767. using the windows calculator, you will get:
3.0518509475997192297128208258309e-5

the E-5 means you have you have to move the decimal point 5 places to the left, and this leaves you with:
0.000030518509475997192297128208258309

you only need the first 8 digits (excluding the decimal point) of the number, and that is the number that you use to multiply the value-(half of max value) with.

Back to top

Profile PM Email Website

revenant chaos

DBPro Master

18

Years of Service

User Offline

Joined: 21st Mar 2007

Location: Robbinsdale, MN

Posted: 22nd Sep 2007 23:10 Edited at: 22nd Sep 2007 23:10

Link

This wouldnt happen to be for a PC Elements recoil game pad is it?

I just thought of an easier way that requires no math to be done

+ Code Snippet

converted=((currentvalue-32767.0)/32767.0)

the same demo from above, but using the new method:

+ Code Snippet

currentvalue=32767
converted as float

do
cls
set cursor 0,0
print "current value=",currentvalue
converted=((currentvalue-32767)/32767.0)
print "the converted value=",converted

currentvalue=currentvalue+((upkey()-downkey())*10)
if currentvalue<0 then currentvalue=0
if currentvalue>65535 then currentvalue=65535

loop

the "converted" variable needs to be a float, and the number that it is divided by (the numerator) must be expressed as a float, otherwise it will return an integer value. -1 at 0, 0 at 32767, and 1 at 65535.

Back to top

Profile PM Email Website

IanM

Retired Moderator

22

Years of Service

User Offline

Joined: 11th Sep 2002

Location: In my moon base

Posted: 22nd Sep 2007 23:12

Link

Ideally, you should take the first 8 significant digits rather than 8 decimal places to avoid too much loss of precision. i.e. in this example, I'd use 0.000030518509, rounding the last digit as appropriate according to the next digit.

It won't make too much difference on the values being used here (+-32767 to +-100), but for scaling upwards especially for larger ratios, the difference may be quite noticeable.

Utility plugins collection and
http://www.matrix1.demon.co.uk for older plug-ins and example code

Back to top

Profile PM Email Website

Green Gandalf

VIP Member

20

Years of Service

User Offline

Joined: 3rd Jan 2005

Playing: Malevolence:Sword of Ahkranox, Skyrim, Civ6.

Posted: 22nd Sep 2007 23:17

Link

Of course, neither of those solutions work correctly (and the original question wasn't quite right either - what's half of 65535?).

Back to top

Profile PM Email

kbyoyoa

18

Years of Service

User Offline

Joined: 30th Apr 2007

Location: Minnesota

Posted: 22nd Sep 2007 23:39 Edited at: 22nd Sep 2007 23:40

Link

ok I work on this for a few minutes and came up with this function...

+ Code Snippet

function convertrange(in,oldlow,oldhigh,newlow,newhigh)
rangeold# = oldhigh-oldlow
rangenew# = newhigh-newlow
ratio# = rangeold#/rangenew#
out = (newlow + (in-oldlow)/ratio#)
endfunction out

this will convert the number "in" in the range "oldlow" to "oldhigh" to a number in the same position in the range "newlow" to "newhigh"

This will work for any range not just the one you stated.

hope this is what you where looking for...

Back to top

Profile PM Email Website

IanM

Retired Moderator

22

Years of Service

User Offline

Joined: 11th Sep 2002

Location: In my moon base

Posted: 23rd Sep 2007 00:12

Link

Quote: "Of course, neither of those solutions work correctly (and the original question wasn't quite right either - what's half of 65535?)."

You're right about the solution being wrong, as it concentrates on start/end points, and misses the fact that we're talking about inclusive ranges. -100 to 100 has 201 integers in it, and 0 to 65535 has 65536 integers in it - (endpoint - startpoint + 1).

Here's working code, that's not quite right yet:

+ Code Snippet

print Adjust(    0, 0, 65535, -100, 100)
print Adjust(32000, 0, 65535, -100, 100)
print Adjust(32767, 0, 65535, -100, 100)
print Adjust(33000, 0, 65535, -100, 100)
print Adjust(65535, 0, 65535, -100, 100)

wait key
end


function Adjust(Value as integer, LowIn as integer, HighIn as integer, LowOut as integer, HighOut as integer)
   local InRange as integer
   local OutRange as integer
   local Ratio as float

   InRange = (HighIn - LowIn) + 1
   OutRange = (HighOut - LowOut) + 1

   Ratio = (OutRange + 0.0) / InRange

   Value = ((Value + LowIn) * Ratio) + LowOut
endfunction Value

The problem is that the rounding isn't quite right just yet - it needs to round to the nearest integer rather than truncating as it does now

Utility plugins collection and
http://www.matrix1.demon.co.uk for older plug-ins and example code

Back to top

Profile PM Email Website

Sorry your browser is not supported!

DarkBASIC Professional Discussion / Converting number ranges