Prev: Excellent Oracle Apps Technical Consutant avaliable immediately for your client requriments
Next: Dvalve! we called the heart of outsourcing world.
From: david jensen on 9 Apr 2010 16:10
Hi all,

I'm trying to assign a color to various values between some arbitrary
maximum and minimum, and i'd like to go from red to blue to green .
Currently, I have the following, where i'm passing the minimum and
maximum as a 2-tuple:

def getcolor( minmax, curr ):
rangesize = (minmax[1] - minmax[0]) - 1
currsize = float( curr - minmax[0] )
currpos = currsize / rangesize
colornum = int( currpos*511 + 0.5 )
r = max( 255 - colornum, 0 )
g = max( -256 + colornum, 0 )
b = 255 - ( r+g )
return r,g,b

1) is there a better way to do this?
2) at the extreme points, this doesn't work: the number of (255, 0,
0)s and (0, 255, 0)s are 1/2 what they should be, and (0, 0, 255) is
double what it should be.

the above algorithm at least doesn't give me things like (0, -1, 256),
which i struggled with for a while, but where is/are my off-by-one(s)?

many thanks

d
From: Terry Reedy on 9 Apr 2010 19:41
On 4/9/2010 4:10 PM, david jensen wrote:
> Hi all,
>
> I'm trying to assign a color to various values between some arbitrary
> maximum and minimum, and i'd like to go from red to blue to green .
> Currently, I have the following, where i'm passing the minimum and
> maximum as a 2-tuple:
>
> def getcolor( minmax, curr ):

One would need more info re minmax and curr and their relationship to
definitively answer your question.

> rangesize = (minmax[1] - minmax[0]) - 1

I would expect that either +0 or +1 woult be correct

> currsize = float( curr - minmax[0] )
> currpos = currsize / rangesize
> colornum = int( currpos*511 + 0.5 )

The .5 may be the reason for the '1/2' question.

> r = max( 255 - colornum, 0 )
> g = max( -256 + colornum, 0 )

r,g are both 0 for both 255 and 256. I suspect this is reason for the
'double' question.

> b = 255 - ( r+g )
> return r,g,b
>
> 1) is there a better way to do this?
> 2) at the extreme points, this doesn't work: the number of (255, 0,
> 0)s and (0, 255, 0)s are 1/2 what they should be, and (0, 0, 255) is
> double what it should be.
>
> the above algorithm at least doesn't give me things like (0, -1, 256),
> which i struggled with for a while, but where is/are my off-by-one(s)?

Again, start with a complete definition of the params.

Terry Jan Reedy

 | 
Pages: 1
Prev: Excellent Oracle Apps Technical Consutant avaliable immediately for your client requriments
Next: Dvalve! we called the heart of outsourcing world.