Converting negative integers to HEX values
Arun Chhabra
To respond to this message, remove "no-spam" from the reply-to address
*****
I was trying to convert decimal integer values to HEX format. As per the
"dec2hex" command it appears that this should not be a problem.
However, I am not able to convert a negative number to the expected HEX
value (or vice versa using the hex2dec command). As I understand, this
command is meant to work with "integer decimal" values (or convert to it, in
the case of dec2hex).
Specifically, as an example, consider the decimal value:
-5698. It's equivalent hex value should be FFFFE9C4, instead Matlab 4.0
gives:
-5.698000e+003.
OR
-10 should be FFFFFFF6, but instead the result output by Matlab is:
-1.000000e+001
I would appreciate any help that you are able to provide.
Regards,
Arun
Alexandre Kampouris
Arun Chhabra wrote:
> However, I am not able to convert a negative number to the expected HEX value
> (or vice versa using the hex2dec command). As I understand, this command is
> meant to work with "integer decimal" values (or convert to it, in the case of
> dec2hex).
> Specifically, as an example, consider the decimal value:
>
> -5698. It's equivalent hex value should be FFFFE9C4, instead Matlab 4.0
> gives:
> -5.698000e+003.
On Matlab 5, this returns an error :
» dec2hex(-5,3)
??? Error using ==> dec2hex
D must be a non-negative integer.
For integers there are several ways to represent negative numbers.
I'd suppose you're looking for two's complement representation, which is the
most common:
function [s]=myhex(i,n);
if i<0,
s=dec2hex(16^n+i,n);
else
s=dec2hex(i,n);
end;
This can be improved to meet your particular needs... There is no protection or
error checking in this version. I leave it as an exercise to modify this for an
arbitrary word length, instead of a number of 4 bit nibbles. :-) And remember
that Matlab works with values in IEEE floating point representation, with a
mantissa of 53 bits or so. Don't try converting too wide a field.
The two's complement representation of a negative number will depend on the word
length, whereas a positive number will always be the same, if you ignore the
varying number of leading zeros.
Alexandre
Troy Richards
Jeremy,
Check out www.nr.com and look at on-line book in chapter 4.5 on Gaussian
Quadratues. I have converted gaulag and gauher from fortran to matlab if
you are interested.
--
Troy Richards
e-mail troy.r...@drea.dnd.ca
Phone (902)426-3100 Ext 285
FAX (902)426-9654
---------------------------------
Mail Delivery
DREA
PO BOX 1012
Dartmouth NS, B2Y 3Z7
---------------------------------
Courier Delivery
DREA
9 Grove St.
Dartmouth NS, B3A 3C5
---------------------------------