| Re: SIGN function: signof(num or array of num) [message #70131] |
Tue, 16 March 2010 09:41 |
Foldy Lajos
Messages: 268
Registered: October 2001
|
Senior Member |
|
|
On Tue, 16 Mar 2010, Ding wrote:
> although signof(0) cause algorithm error, since 0/0 is defined as 0,
> signof(0)=0, signof(Inf)=0. it can be used in an expression
> signof(numeric variable or array)*10.0/5. etc.
0/0 is undefined in mathematics.
regards,
lajos
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70132 is a reply to message #70131] |
Tue, 16 March 2010 09:36  |
Carsten Lechte
Messages: 124
Registered: August 2006
|
Senior Member |
|
|
I wrote:
> One would have to extract the sign bit (assuming IEEE format,
> taking into account endianness, precision etc.) and base the
> calculation on that.
Alternatively, one could read the documentation and discover the
SIGN keyword to the FINITE() function ;-)
chl
|
|
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70134 is a reply to message #70133] |
Tue, 16 March 2010 09:17 |
Ding
Messages: 20
Registered: March 2008
|
Junior Member |
|
|
On Mar 16, 3:46 pm, Carsten Lechte <c...@toppoint.de> wrote:
> Ding wrote:
>> IDL> a=[[1,0][-1,0]]
>> IDL> print,signof(a)
>> 1 0
>> -1 0
>
> Still, you should mention under RESTRICTIONS that the
> input must be an integer type:
>
> IDL> a=[[1,0],[-1,0]]
> IDL> print, signof(a)
> 1 0
> -1 0
> % Program caused arithmetic error: Integer divide by 0
> IDL> print, signof(double(a))
> 1 -32768
> -1 -32768
> % Program caused arithmetic error: Floating illegal operand
> IDL> print, signof(complex(a))
> 1 -32768
> -1 -32768
> % Program caused arithmetic error: Floating illegal operand
>
> chl
This is all the information I have, including the !version:
IDL> print,!version
{ x86_64 linux unix linux 6.4 Apr 26 2007 64 64}
IDL> print,a
1 0
-1 0
IDL> print,double(a)
1.0000000 0.0000000
-1.0000000 0.0000000
IDL> print,complex(a)
( 1.00000, 0.00000)( 0.00000, 0.00000)
( -1.00000, 0.00000)( 0.00000, 0.00000)
IDL> print,signof(a)
1 0
-1 0
% Program caused arithmetic error: Integer divide by 0
IDL> print,signof(double(a))
1 0
-1 0
% Program caused arithmetic error: Floating illegal operand
IDL> print,signof(complex(a))
1 0
-1 0
% Program caused arithmetic error: Floating illegal operand
IDL>
Let's compare the version of the result and see if it is IDL or
processor that cause the difference!
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70136 is a reply to message #70134] |
Tue, 16 March 2010 08:46 |
Carsten Lechte
Messages: 124
Registered: August 2006
|
Senior Member |
|
|
Ding wrote:
> IDL> a=[[1,0][-1,0]]
> IDL> print,signof(a)
> 1 0
> -1 0
Still, you should mention under RESTRICTIONS that the
input must be an integer type:
IDL> a=[[1,0],[-1,0]]
IDL> print, signof(a)
1 0
-1 0
% Program caused arithmetic error: Integer divide by 0
IDL> print, signof(double(a))
1 -32768
-1 -32768
% Program caused arithmetic error: Floating illegal operand
IDL> print, signof(complex(a))
1 -32768
-1 -32768
% Program caused arithmetic error: Floating illegal operand
chl
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70137 is a reply to message #70136] |
Tue, 16 March 2010 08:43 |
Ding
Messages: 20
Registered: March 2008
|
Junior Member |
|
|
On Mar 16, 10:51 am, Ding <gardener_2...@hotmail.com> wrote:
> Dear IDL users,
>
> I cannot find the SIGN function in IDL procedures, but one in
> Solarsoft ssw/maths/ SIGN which occupy the name, but different
> purpose( sign(a)*b). so I wrote the SIGNOF() function which is to get
> the sign of a number or array,
>
> Hopefully I am not repeating someone' work! you are welcome to
> comments!
>
> ; return the sign of the input data in an array of the same size
> ; -1 for negative elements
> ; 1 for positive elements
> ; 0 for 0 and complex number
> ; only works for numerics
return the sign of the input data in an array of the same size
-1 for negative elements
1 for positive elements
0 for 0 and complex number
only works for numerics
>a=[[1,-2],[0,3]]
> print, signof(a)
1 -1
0 1
>
More notes:
although signof(0) cause algorithm error, since 0/0 is defined as 0,
signof(0)=0, signof(Inf)=0. it can be used in an expression
signof(numeric variable or array)*10.0/5. etc.
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70138 is a reply to message #70137] |
Tue, 16 March 2010 08:35 |
Carsten Lechte
Messages: 124
Registered: August 2006
|
Senior Member |
|
|
H. Evans wrote:
> Could be either + or -ve:
> IDL> print,!values.f_nan, - !values.f_nan
> NaN -NaN
Then, a totally different approach might be necessary to get
it completely right.
FINITE() tells you if you have a NaN, but since every
comparison that involves NaNs evaluates to FALSE, you cannot
find out if it's + or -NaN be equating to !VALUES.F_NAN.
One would have to extract the sign bit (assuming IEEE format,
taking into account endianness, precision etc.) and base the
calculation on that.
For my private version, I will add a term + ( FINITE( x) EQ 0)
so all NaNs are assumed positive, since the one value that
SIGNUM( NaN) should not have is 0.
I wonder in what situation one would care about the correct sign
of something that is not a number.
chl
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70139 is a reply to message #70138] |
Tue, 16 March 2010 08:32 |
Ding
Messages: 20
Registered: March 2008
|
Junior Member |
|
|
On Mar 16, 1:31 pm, FÖLDY Lajos <fo...@rmki.kfki.hu> wrote:
> On Tue, 16 Mar 2010, Ding wrote:
>> Dear IDL users,
>
>> I cannot find the SIGN function in IDL procedures, but one in
>> Solarsoft ssw/maths/ SIGN which occupy the name, but different
>> purpose( sign(a)*b). so I wrote the SIGNOF() function which is to get
>> the sign of a number or array,
>
>> Hopefully I am not repeating someone' work! you are welcome to
>> comments!
>
>> return, fix(abs(input)/input)
>
> IDL> input=0 & help, fix(abs(input)/input)
> <Expression> INT = 0
> % Program caused arithmetic error: Integer divide by 0
>
> I think SIGNOF() should not cause an arithmetic error for finite numbers.
>
> regards,
> lajos
This is expected in my algorithm, as the sign of 0 is considered to be
0. although there is an error in the IDL language, however 0/0=0, is
still true. the power of this in array calculation
IDL> a=[[1,0][-1,0]]
IDL> print,signof(a)
1 0
-1 0
Ding Yuan
|
|
|
|
| Re: SIGN function: signof(num or array of num) [message #70140 is a reply to message #70139] |
Tue, 16 March 2010 07:30 |
H. Evans
Messages: 18
Registered: December 2009
|
Junior Member |
|
|
On Mar 16, 2:57 pm, Carsten Lechte <c...@toppoint.de> wrote:
> F LDY Lajos wrote:
>> I think SIGNOF() should not cause an arithmetic error for finite numbers.
>
> I use the expression
>
> (x GT 0) -1* (x LT 0)
>
> to get the sign of e.g.
>
> x = [ -2e4, 5, 0, 1d/0, -1d/0, 0d/0 ]
>
> but it uses 2 comparisons and even a multiplication. Does anyone have
> a simpler/faster version? Furthermore, what should the sign of NaN be?
>
> chl
Could be either + or -ve:
IDL> print,!values.f_nan, - !values.f_nan
NaN -NaN
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
