r****y
发帖数: 1437 | 1
The simplest way is to just polyfit the log value la.
logcoeff=polyfit(x, log10(y), n)
I remeber there is a subroutine in Matlab can do any formula
fitting, cannot remeber the exactly name of it.
Double check when you use Matlab to do such manipulation. Matlab
is not as reliable as it is declared. I was one more experience about it
in recent. | q***o
发帖数: 484 | 2
try to use
lsqcurvefit, which is more powerful and also more complex.
【在 r****y 的大作中提到】
:
: The simplest way is to just polyfit the log value la.
: logcoeff=polyfit(x, log10(y), n)
: I remeber there is a subroutine in Matlab can do any formula
: fitting, cannot remeber the exactly name of it.
: Double check when you use Matlab to do such manipulation. Matlab
: is not as reliable as it is declared. I was one more experience about it
: in recent.
|
|
