Re: Distinguish between data and noise
- From: Olin Perry Norton <norton@xxxxxxxxxxxxxxxxxxxx>
- Date: Wed, 03 Jan 2007 09:13:44 -0600
Alexander Müller wrote:
Hello!
I want to analyze spectra (count rates/voltage over wavelength).
How can I calculate the noise level in a given intervall, for example
over 10 Points? Or in other words: how can I distinguish between
a peak and random noise?
Alex
Alex,
If you are counting something, such as with a particle
detector, photomultiplier tube, etc.) one source of noise
is due to the random nature of the particle arrival times.
You should read about Poisson statistics.
There is some discussion about fitting peaks in spectra
in Lyons, "Statistics for nuclear and particle physicists,"
and in Bevington and Robinson, "Data reduction and data
analysis for the physical sciences."
Olin Perry Norton
.
- Prev by Date: Re: numerical analysis (composite numerical integration)
- Next by Date: Re: Looking for suggestions on a root search strategy
- Previous by thread: Re: numerical analysis (composite numerical integration)
- Next by thread: Multiplication in a "band matrix" space
- Index(es):
Relevant Pages
|
