Release Notes Version 6 0 and Poissonscale: Difference between pages

From Eigenvector Research Documentation Wiki
(Difference between pages)
Jump to navigation Jump to search
imported>Jeremy
 
imported>Jeremy
(Created page with "===Purpose=== Perform Poisson scaling with scaling offset. ===Synopsis=== :[xs,sc] = poissonscale(x,''options''); %calibrate scaling :xs = poissonscale(x,sc); ...")
 
Line 1: Line 1:
Version 6.0 of PLS_Toolbox and Solo was released in October, 2010.
===Purpose===


For general product information, see [http://www.eigenvector.com/software/pls_toolbox.htm PLS_Toolbox Product Page]. For information on Solo, see [http://www.eigenvector.com/software/solo.htm Solo Product Page].
Perform Poisson scaling with scaling offset.


(back to [[Release Notes PLS Toolbox and Solo]])
===Synopsis===


==NEW FEATURES==
:[xs,sc] = poissonscale(x,''options'');    %calibrate scaling
:xs = poissonscale(x,sc);              %apply previous scaling


===Model Building===
===Description===
* [[Image:Clutterbutton.png|right]]New "top-level" access to [[Declutter_Settings_Window|Decluttering methods]] (GLSW, EPO, EMM) through "Clutter" control in Analysis window
* Revised [[WorkspaceBrowser_Layout|Workspace Browser]] with new features and improved look
* Automatic model report generator [http://www.eigenvector.com/eigenguide.php?m=Report_Writer (watch how-to movie) ]  [ [[reportwriter]] ]
* Orthogonalized PLS models (compare to O-PLS) [ [[orthogonalizepls]] ]
* NPLS (multiway PLS) now available in Analysis window [ [[Analysis_Menu]] ]
* Non-linear modeling with polynomial transformations [ [[polytransform]] ]
* New "contrast" option in MCR-ALS for feasible bounds results [ [[MCR Contrast Constraint]] ]
* Automatic calibration subset selection tool [ [[reducennsamples]] ]
* RMSEP reported as a function of LVs
* Scripting language and interface for user scripting and 3rd party product control [ [[Solo Predictor Script Construction]] ]


===Ease-of-Use Features===
Scales each variable by its square root mean value. When no scale values
*  Customizable class symbols in plots [http://www.eigenvector.com/eigenguide.php?m=Modifying_Class_Symbols (watch how-to movie)] [ [[Set Symbol Styles Window]] ]
are passed, a calibration is performed in which the square root mean
* Allow plotting of loadings in original units (with preprocessing undone) [[Image:undoprepro.png]]
values are calculated for each variable and then these are applied to the
* Regression error bars can be turned on and off [[Image:showerrorbars.png]]
input data. If previously calculated scales are passed, these are simply
* Automatic augmentation of calibration set from validation data
applied to the data. An optional options structure allows setting of the
* Clustering - class and label information available in "mouse-over" of dendrogram
offset which is added to each mean to avoid over-scaling when a variable
* New interface and object for database queries
has a near-zero mean.


===Functionality Improvements===
* Improved SVM performance
* Improved model cache performance
* Improved automated initial guesses for MCR
* Better compatibility between boxplot and Mathworks-related functions
* Streamlined interface menus
* User-adjustable pane sizes in primary interfaces
* Improved model and prediction information in Analysis window
* SPC reader performance improved
* XML writer - significant speed improvements


===Documentation===
====Inputs====
* Significant documentation enhancements including new [[Software_User_Guide|"getting started"]] guide and [http://www.eigenvector.com/eigenguide.php how-to movies]


===PLS_Toolbox-Specific Enhancements===
* '''x''' = Data to be scaled (double or DataSet object).
* Improved PLS_Toolbox installer
* Varimax rotations on standard model structures (PLS_Toolbox only)
* Crossval now accepts model as input (automatic population of results)
* Cluster -allow access to raw sample-to-sample distance matrix (distances)
* Exteriorpts -improved performance


==NEW FUNCTIONS AND FILES==
====Optional Inputs====


:[[cov_cv]] - Estimation of a regularized inverse covariance matrix.
* '''sc''' = Vector of previously-calculated scales. Must be equal in length to the number of included x-block variables.
:[[evridir]] - Locate and or create EVRI home directory.
* '''options''' = Options structure. See Options section below.
:[[polytransform]] - Add polynomial and cross terms to data matrix.
 
:[[reducennsamples]] - Select a subset of samples by removing nearest neighbors.
====Outputs====
:[[reportwriter]] - Write a summary of the analysis including associated figures to html/word/powerpoint.
 
:[[symbolstyle]] - Interface to modify class symbols.
* '''xs''' = Scaled data.
* '''sc''' = Vector of scales calculated for given data.
 
===Options===
 
''options'' =  a structure array with the following fields:
 
* '''offset''': [ 3] percent of the maximum mean value to be used as an offset on all scales. Avoids division by near-zero means.
 
===See Also===
 
[[auto]], [[preprocess]], [[rescale]], [[scale]]

Revision as of 15:46, 26 July 2011

Purpose

Perform Poisson scaling with scaling offset.

Synopsis

[xs,sc] = poissonscale(x,options); %calibrate scaling
xs = poissonscale(x,sc); %apply previous scaling

Description

Scales each variable by its square root mean value. When no scale values are passed, a calibration is performed in which the square root mean values are calculated for each variable and then these are applied to the input data. If previously calculated scales are passed, these are simply applied to the data. An optional options structure allows setting of the offset which is added to each mean to avoid over-scaling when a variable has a near-zero mean.


Inputs

  • x = Data to be scaled (double or DataSet object).

Optional Inputs

  • sc = Vector of previously-calculated scales. Must be equal in length to the number of included x-block variables.
  • options = Options structure. See Options section below.

Outputs

  • xs = Scaled data.
  • sc = Vector of scales calculated for given data.

Options

options = a structure array with the following fields:

  • offset: [ 3] percent of the maximum mean value to be used as an offset on all scales. Avoids division by near-zero means.

See Also

auto, preprocess, rescale, scale

Retrieved from "https://www.wiki.eigenvector.com/index.php?title=Release_Notes_Version_6_0&oldid=7604"