I have data (stress-strain) from tension-compression test. But the data have noise at some data points. How can I eliminate the data points which do not represent the corresponding behavior?
First I'd make sure your "noise" isn't reality. For example, a composite might experience a large number of fiber failures starting early in the test, and the release of that energy might look pretty noisy.
But if you're sure it's not the case, my next look would be at an anti-aliasing filter.
https://en.wikipedia.org/wiki/Anti-aliasing_filter
This will ensure that any noise above the Nyquist frequency isn't being coupled on and appearing as data.
First I'd make sure your "noise" isn't reality. For example, a composite might experience a large number of fiber failures starting early in the test, and the release of that energy might look pretty noisy.
But if you're sure it's not the case, my next look would be at an anti-aliasing filter.
https://en.wikipedia.org/wiki/Anti-aliasing_filter
This will ensure that any noise above the Nyquist frequency isn't being coupled on and appearing as data.
Hi Eray
Since the stress-strain data are experimentally obtained,you can use a non-linear regression for fitting a model to the experimental data, regardless of the presence of noise at some data points.
Please Refer to ;
-H. Motulsky, and A. Christopoulos, Fitting Models to Biological Data Using Linear and Nonlinear Regression: A practical guide for Curve Fitting, GraphPad Software Inc., San Diego, CA(2003).www.graphpad.com.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction
- XRD Analysis