PLS pridiction error in Unscrambler Log Out | Topics | Search
Moderators | Register | Edit Profile

NIR Discussion Forum » Bruce Campbell's List » I need help » PLS pridiction error in Unscrambler « Previous Next »

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

venkatarman (venkynir)
Senior Member
Username: venkynir

Post Number: 168
Registered: 3-2004
Posted on Saturday, March 30, 2013 - 11:11 am:   

Hi Velu :
Welcome to NIR discussion group . Have under gone training at M/s Camo India .If not please do that . Definitely you will get dimension error .Usually your training set dimension should match with prediction set.Please see Video of Unscrambler.
Top of pagePrevious messageNext messageBottom of page Link to this message

C.Kumaravelu (ceekeyvelu)
New member
Username: ceekeyvelu

Post Number: 1
Registered: 3-2013
Posted on Saturday, March 30, 2013 - 9:18 am:   

Dear all,
I am C.Kumaravelu, Principal Scientist, CSIR-CEERI,Chennai, India, is an R&D lab. I am working in the area of Agro-food industry application using NIR spectroscopy and chemometrics. My problem is I am unable to predict my test samples using the developed calibrated PLS model. I did the model using CAMO software Unscrambler with the the data set (X) by 114X2201 (No. of samples X absorption at wavelength from 1100nm to 2100nm) as a one matrix set. Also I added in the 2202nd column as the Discriminate Analysis class and in 2203rd column as its logical value (Y) - (either 0 or 1). Now I have a data set 114X2203, for which I did the PLS calibration model. In order to predict the class of new sample (either 0 or 1), I fed the test sample data set of 34X2201, I am unable to use this data set to predict and getting an error is data dimension error. Could someone help that without DA class and Logical value of test samples, how can I predict the 34 X2201 data set using the 114X2203 PLS calibrated model.

Thank you
Regards
C.Kumaravelu

Add Your Message Here
Posting is currently disabled in this topic. Contact your discussion moderator for more information.