'Code Break' Thread RSS Feed

Code Break

Mon, 25 Feb 2008 19:51:54 -0700

My code breaks at this point <in red>. Any suggestions?

int compare (const void * a, const void * b)
{
return ( *(float*)a - *(float*)b );
}

float Calc_Median(int intArraySize)
{
int n;
qsort (CalcTimeArray, intArraySize, sizeof(float), compare);
for (n=0; n<intArraySize; n++)
printf ("%d ",CalcTimeArray[n]);
return 0;
}

Re: Code Break

Mon, 25 Feb 2008 20:50:03 -0700

: My code breaks at this point <in red>. Any suggestions?
:
: int compare (const void * a, const void * b)
: {
: return ( *(float*)a - *(float*)b );
: }
:
: float Calc_Median(int intArraySize)
: {
: int n;
: qsort (CalcTimeArray, intArraySize, sizeof(float), compare);
: for (n=0; n<intArraySize; n++)
: printf ("%d ",CalcTimeArray[n]);
: return 0;
: }
:

What's "code breaks"?
1. Code does not compile
2. Code does not run

Re: Code Break

Tue, 26 Feb 2008 06:56:35 -0700

: : My code breaks at this point <in red>. Any suggestions?
: :
: : int compare (const void * a, const void * b)
: : {
: : return ( *(float*)a - *(float*)b );
: : }
: :
: : float Calc_Median(int intArraySize)
: : {
: : int n;
: : qsort (CalcTimeArray, intArraySize, sizeof(float), compare);
: : for (n=0; n<intArraySize; n++)
: : printf ("%d ",CalcTimeArray[n]);
: : return 0;
: : }
: :
:
: What's "code breaks"?
: 1. Code does not compile
: 2. Code does not run
:

Code doesn't run. I put break points and stepped thru. Then the code does do anything when it gets to the return (* (float* ........

Re: Code Break

Tue, 26 Feb 2008 09:03:36 -0700

Since you didn't post what type of variables you put into qsort, the question is impossible to answer. There is no error in that code. (Except it is non-portable, but that might not be an issue here)

Re: Code Break

Tue, 26 Feb 2008 11:16:07 -0700

: Since you didn't post what type of variables you put into qsort, the
: question is impossible to answer. There is no error in that code.
: (Except it is non-portable, but that might not be an issue here)

Figured out that it didn't like using "float" but did fine with the "int". I went to using a vector class. Works the way I want it to.

Thanks....

Re: Code Break

Thu, 28 Feb 2008 21:14:48 -0700

: : Since you didn't post what type of variables you put into qsort, the
: : question is impossible to answer. There is no error in that code.
: : (Except it is non-portable, but that might not be an issue here)
:
: Figured out that it didn't like using "float" but did fine with the
: "int". I went to using a vector class. Works the way I want it to.
:
: Thanks....
:
:
:
:
:

Hmm, on that note I thought I would contribute a reason. Type casting it to float means that it would use floating point arithmatic. Which simply doesn't make sense to use in that context. For example thinking of floating point arithmatic in terms of scientific notation, they are very similar:

if you go

1.45E-1 + 4.00E-1 you get 5.45E-1 but if you add 1.34E0 and 5.00E-1 you get 1.84E0 which is represented you completely different in terms of binary than the sum of each direct number.

Aye aye aye.