| Re: more bugs in envi !!! [message #35811 is a reply to message #35725] |
Mon, 14 July 2003 13:07   |
gauravjn123
Messages: 7
Registered: July 2003
|
Junior Member |
|
|
Hi all!
Thanx for the help...well u both were right...i was under the wrong
impression that i had defined it as a long integer...but i hadn't. But
i still cant find out why these 2 functions are producing different
outputs :
Please go through the 2 functions:
>>
>> 1.function example,b1
>> help,b1
>> end
>> 2.function example,b1,b2
>> help,b1
>> end
>>
>> Now could anyone tell me why these 2 codes are producing different
>> final answers??? I just cant find any reason for it.
Gaurav Jain
ENST-Bretagne
Marc Schellens <m_schellens@hotmail.com> wrote in message news:<3F115ED4.6040200@hotmail.com>...
>> Hi folks!
>> Sorry to trouble u again....but envi seems to be just going above my
>> head...there is another problem that i am facing:
>> Please go through the 2 functions:
>>
>> 1.function example,b1
>> help,b1
>> end
>> 2.function example,b1,b2
>> help,b1
>> end
>>
>> Now could anyone tell me why these 2 codes are producing different
>> final answers??? I just cant find any reason for it.
>>
>> One more problem :
>>
>> 'b1' is a positive array and 'avgb1' is a positive number but when i
>> use the formula : result=b1*100/avgb1
>> print, result
>> , the answer that i get is a negative array.
>>
>> Whereas if i use the formula : result=b1/avgb1
>> print, result * 100
>> , then i get a positive array
>>
>
> Even without knowing ENVI, the second problem isn't a bug in ENVI.
>
> A integer occupies (in IDL) 16-bit, and therfore covers a range form
> -32768 to
> 32767
>
> anything larger (or smaller) cannot be represented and therefore you
> observe an 'overflow', resulting in a negative number.
> IDL evalutates your expressinon from left to right, the
> overflow occurs after the multiplication with 100.
> In the second case the division is done first and thus the value seems
> to be small enough not to overflow in the later multiplication.
>
> The solution is to convert your b1 array to LONG.
> But did you notice that you do an integer division?
> 1/2 equals 0.
> Maybe you want to use FLOAT or DOUBLE.
>
> hdh,
> marc
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
