Figmerit: Difference between revisions

From Eigenvector Research Documentation Wiki
Jump to navigation Jump to search
imported>Jeremy
(Importing text file)
imported>Bob
No edit summary
Line 1: Line 1:
===Purpose===
===Purpose===


Line 10: Line 9:
===Description===
===Description===


Calculates analytical figures of merit for PLS and PCR standard model structures. Inputs are the preprocessed (usually centered and scaled) spectral data x, the preprocessed analyte data y, and the regression vector, b. Note that for standard PLS and PCR structures b = model.reg.
Calculates analytical figures of merit for PLS and PCR standard model structures. Inputs are the preprocessed (usually centered and scaled) spectral data <tt>x</tt>, the preprocessed analyte data <tt>y</tt>, and the regression vector, <tt>b</tt>. Note that for standard PLS and PCR structures <tt>b = model.reg</tt>.
 
The outputs are the matrix of net analyte signals <tt>nas</tt> for each row of <tt>x</tt>, the norm of the net analyte signal for each row <tt>nnas</tt> (this is corrected to include the sign of the prediction), the matrix of sensitivities for each sample <tt>sens</tt>, and the vector of selectivities for each sample <tt>sel</tt> (sel is always non-negative).


The outputs are the matrix of net analyte signals nas for each row of x, the norm of the net analyte signal for each row nnas (this is corrected to include the sign of the prediction), the matrix of sensitivities for each sample sens, and the vector of selectivities for each sample sel (sel is always non-negative).
Note that the "noise-filtered" estimate present in previous versions is no longer used because an improved method for calculating the net analyte vector makes it redundant.


Note that the "noise-filtered" estimate present in previous versions is no longer used because an improved method for calculating the net analyte vector makes it redundant
====Inputs====


===Examples===
* '''x''' = x-block data, normally centered and scaled
* '''y''' = y-block data, preprocessed
* '''b''' = regression vector.  Standard PLS_Toolbox PLS and PCR structures contain this vector in the <tt>.reg</tt> field.


Given the 7 LV PLS model:
====Outputs====


:modl = pls(x,y,7);
* '''nas''' = net analyte signals for each row of <tt>x</tt>.
* '''nnas''' = norm of the net analyte signal for each row.
* '''sens''' = matrix of sensitivities for each sample.
* '''sel''' = vector of selectivities for each sample.


:Rhat = modl.loads{1,1}\*modl.loads{2,1}';


:[nas,nnas,sens,sel,nfnas] = figmerit(x,y,Rhat);
===Examples===


Given the 5 PC PCR model:
Given the 7 LV PLS model:


:modl = pcr(auto(x),auto(y),5);
<pre>
modl = pls(x,y,7);
Rhat = modl.loads{1,1}\*modl.loads{2,1}';
[nas,nnas,sens,sel,nfnas] = figmerit(x,y,Rhat);
</pre>


:Rhat = modl.loads{1,1}\*modl.loads{2,1}';
Given the 5 PC PCR model:


:[nas,nnas,sens,sel,nfnas] = figmerit(auto(x),auto(y),Rhat);
<pre>
modl = pcr(auto(x),auto(y),5);
Rhat = modl.loads{1,1}\*modl.loads{2,1}';
[nas,nnas,sens,sel,nfnas] = figmerit(auto(x),auto(y),Rhat);
</pre>


===See Also===
===See Also===


[[pcr]], [[pls]]
[[pcr]], [[pls]]

Revision as of 11:47, 9 October 2008

Purpose

Analytical figures of merit for multivariate calibration.

Synopsis

[nas,nnas,sens,sel] = figmerit(x,y,b);

Description

Calculates analytical figures of merit for PLS and PCR standard model structures. Inputs are the preprocessed (usually centered and scaled) spectral data x, the preprocessed analyte data y, and the regression vector, b. Note that for standard PLS and PCR structures b = model.reg.

The outputs are the matrix of net analyte signals nas for each row of x, the norm of the net analyte signal for each row nnas (this is corrected to include the sign of the prediction), the matrix of sensitivities for each sample sens, and the vector of selectivities for each sample sel (sel is always non-negative).

Note that the "noise-filtered" estimate present in previous versions is no longer used because an improved method for calculating the net analyte vector makes it redundant.

Inputs

  • x = x-block data, normally centered and scaled
  • y = y-block data, preprocessed
  • b = regression vector. Standard PLS_Toolbox PLS and PCR structures contain this vector in the .reg field.

Outputs

  • nas = net analyte signals for each row of x.
  • nnas = norm of the net analyte signal for each row.
  • sens = matrix of sensitivities for each sample.
  • sel = vector of selectivities for each sample.


Examples

Given the 7 LV PLS model:

 modl = pls(x,y,7);
 Rhat = modl.loads{1,1}\*modl.loads{2,1}';
 [nas,nnas,sens,sel,nfnas] = figmerit(x,y,Rhat);

Given the 5 PC PCR model:

 modl = pcr(auto(x),auto(y),5);
 Rhat = modl.loads{1,1}\*modl.loads{2,1}';
 [nas,nnas,sens,sel,nfnas] = figmerit(auto(x),auto(y),Rhat);

See Also

pcr, pls