PCA decomposition of NIR

jlg280's picture


I have run across an interesting signature in my NIR data.  I am collecting an NIR absorbance profile (wavelengths 906 to 1688) on seeds that are dropped down an illuminated light tube.   When I decompose the data using PCA and plot the log of the eigen values, I see a distinct step in the values at value 256.  I have done a similar decomposition of other seed NIR data from another apparatus and see the same step at 256.  I'm hoping that someone can help me identifiy the source of this step.  

Uploaded Images: 
ianm's picture

I suspect that both devices use the same spectrometer and array.

#256 is the last one ..correct?  We always discard the first 5 and the last 5 from the arrays

Dr. Charles R. Hurburgh, Jr.
Professor, Agricultural and Biosystems Engineering Professor in Charge, Iowa Grain Quality Initiative
3167 NSRIC
Iowa State University, Ames. Iowa  50011

(Posted on his behalf by Ian Michael)