Conload: Difference between revisions
Jump to navigation
Jump to search
imported>Jeremy (Importing text file) |
imported>Jeremy (Importing text file) |
||
Line 14: | Line 14: | ||
Note that due to non-orthogonal loadings in PARAFAC, individual correlations can add to more than 1. Therefore, such loadings are not drawn with ellipses but squares added. Use options.force = 'ellipse' or 'square' to force one or the other on the plot. | Note that due to non-orthogonal loadings in PARAFAC, individual correlations can add to more than 1. Therefore, such loadings are not drawn with ellipses but squares added. Use options.force = 'ellipse' or 'square' to force one or the other on the plot. | ||
==== | ====Inputs==== | ||
* '''X''' = modeled data | * '''X''' = modeled data | ||
Line 22: | Line 22: | ||
* '''mode''' = loading mode to investigate (i.e. mode = 1 for samples if they are in the first mode) | * '''mode''' = loading mode to investigate (i.e. mode = 1 for samples if they are in the first mode) | ||
==== | ====Outputs==== | ||
Bcon = Congruence loadings | Bcon = Congruence loadings |
Revision as of 16:32, 3 September 2008
Purpose
Congruence loadings for PARAFAC, TUCKER and NPLS.
Synopsis
- Bcon = conload(X,model,options)
Description
Determines congruence (earlier known as correlation) loadings for a specific mode of a model. Congruence loadings look at "non-average correlations", hence take differences in offset into account.
Note that due to non-orthogonal loadings in PARAFAC, individual correlations can add to more than 1. Therefore, such loadings are not drawn with ellipses but squares added. Use options.force = 'ellipse' or 'square' to force one or the other on the plot.
Inputs
- X = modeled data
- model = standard model structure
- mode = loading mode to investigate (i.e. mode = 1 for samples if they are in the first mode)
Outputs
Bcon = Congruence loadings
Options
- plots : [ 'none' | {'final'} ] Governs the creation of plot of the results.
- force : [ {'off'} | 'ellipse' | 'square' ] Forces a given type of limit on the plots (if plot is given).