[EMAIL PROTECTED] said:
> Compute("a<=0 ? minfloat : log10(a)"),  

[EMAIL PROTECTED] said:
> Compute("a<=0 ? log(b)-1 : log10(a)") 

[EMAIL PROTECTED] said:
> Compute("a<=0 ? maxfloat : a") 

Isn't an alternative method to just mark the invalid data items as
invalid positions? Or have I missed something of the original problem
that was stated?

/Nils

Reply via email to