Question about T^2 and Q Residuals...

For the discussion of math. Duh.

Moderators: gmalivuk, Moderators General, Prelates

Posts: 1
Joined: Thu Feb 23, 2017 4:25 pm UTC

Question about T^2 and Q Residuals...

Postby rezangyal » Thu Feb 23, 2017 4:40 pm UTC

What a doozy, right? :)

First, let me say, I am by no means a chemometrician. I'm a Chemist by schooling and trade, but I'm having some difficulty coming up with a "real-world" explaination for what Q Residuals and T^2 actually "tell you" about a sample set.

My predicament:

I've collected 5 spectra of 3 different liquers (Cointreau, Southern Comfort and Knob Creek) and put them through the Eigenvector Solo ringer. After minimal pre-processing (only doing Mean Centering), if I plot Q Residuals (y-axis) versus T^2 (x-axis) I get the below:

What "real-world" statement can I say about the liquors, based on the plot? Is it correct to say that there is little separation within the model (the T^2 is practically the same for all 3)? Can I say that because there is separation across Q (the y-axis), I can tell the 3 liquors apart from one another?

At the end of the day, what is T^2 and Q actually telling me about this particular group of samples? Thanks for any help!

Return to “Mathematics”

Who is online

Users browsing this forum: Bing [Bot] and 10 guests