[Recalc. vs Exper.][Pred. Vs Exper][Recalc. Residuals][Pred. Residuals][SDEC, R2][SDEP, Q2][SDEC & SDEP][PLS plot][PLS-loadings][PLS-partial weights][PCA-loadings][PCA-scores][X val distribution][X sd distribution]
All the commands in this submenu start a 2D plot application except for the two last, which open a histogram application. Follow these link for a full description of their options and commands.
Notice that this is a "tear-off menu", if you click on the dotted line on top of the menu you will transform the submenu in an independent window which will remain open until you specifically close it. This allows a fast access to the plots and is particularly usefull when you plan to open many plots at the same time.
Plot>>>2D plot>>>Recalc. vs. Exper.
Plot>>>2D plot>>>Pred. vs. Exper.
Plot>>>2D plot>>>Recalc. Residuals
Plot>>>2D plot>>>Pred. Residuals
These commands produce 2D plots of the recalculated or predicted Y-values. As the data depends upon the dimensionality of the model, the User is first asked for the dimensionality of the PLS model to represent.
Press the right arrow button to increase the dimensionality or the left arrow button to decrease the dimensionality . When the desired number of components appears in the window press the OK button, or press the Cancel button to abort the operation.
Recalc. vs. Exper.
Recalculated Y-values (vertical axis) are plotted against Experimental Y-values (horizontal axis). Recalculated values are obtained from the latest PLS model at the given dimensionality.
If the PLS model has ever been used to produce external predictions using the command Utilities>>PLS-predictions, the dialog will show a control asking if we want to show the result of the predictions in the plot. If this control is selected, the plot will contains some additional points representing the objects in the external set, usually colored in red, at the possitions corresponding to their experimental an predicted Y.
Pred. vs. Exper.
Predicted Y-values (vertical axis) are plotted against Experimental Y-values (horizontal axis). Predicted values are obtained from the reduced models of the latest PLS model validation at the given dimensionality.
Recalc. Residuals
Recalculated Y residual values (vertical axis) are plotted against Experimental Y-values (horizontal axis). Recalculated residual values are obtained from the latest PLS model at the given dimensionality.
Pred. Residuals
Predicted Y residuals values (vertical axis) are plotted against Experimental Y-values (horizontal axis). Predicted residual values are obtained from the reduced models of latest PLS model validation at the given dimensionality.
Plot>>>2D plot>>>SDEP, Q2
Plot>>>2D plot>>>SDEC & SDEP
SDEC, R2
Two separate 2D plots are presented. The first represents the values of SDEC (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). The second represents the values of R2 (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). Theoretically optimum SDEC (0.000) and R2 (1.000) values are included as yellow lines in the plots.
SDEP, Q2
Two separate 2D plots are presented. The first represents the values of SDEP (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). The second represents the values of Q2 (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). Theoretically optimum SDEP (0.000) and Q2 (1.000) values are included as yellow lines in the plots.
SDEC & SDEP
Two separate 2D plots are presented. The first represents at the same time the values of SDEC and SDEP (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). The second represents at the same time the values of R2 and Q2 (vertical axis) for each of the PLS model dimensionalities tested (horizontal axis). Theoretically optimum SDEC/SDEP (0.000) and R2 /Q2 (1.000) values are included as yellow lines in the plots.
IMPORTANT: We strongly recommend the User to inspect always this plot, after generating any PLS model.
The command presents a scatter plot of any desired component (a) of the X-scores (ta) versus the same component of the Y-scores (ua). This plot shows the inner relationship between the X's and the Y's, and is the best plot to discover non-linearities, influential groups and outliers in the model (see background).
First, the User is prompted to choose the component of both T and U scores to be shown in the plot.
When the dialog window shows the desired settings press the OK button and the 2D plot will be displayed. Press the Cancel button to abort the command.
The plot will represent objects (molecules) in a scatter plot in which the horizontal axis represents the X-scores (T) and the vertical axis represents the Y-scores (U). The inner relationship is shown as a diagonal yellow line.
Plot>>>2D plot>>>PLS-partial weights
Plot>>>2D plot>>>PCA-loadings
Plot>>>2D plot>>>PCA-scores
First, the User is prompted to choose the components to be shown in the horizontal and vertical axis.
In the cases in which variable vectors are represented (PLS-loadings, PLS-partial weights or PCA-loadings) it is possible to highlight single variable blocks by plotting in red the symbol for the variables of this block. The number which appear by the side of the check boxes corresponds to the number of the variable block defined in File>>>Type of variables>>>Modify.
When the dialog window shows the desired settings press the OK button and the 2D plot will be displayed. Press the CANCEL button to abort the command.
PLS-loadings
PLS-partial weights
PCA-loadings
Plot variables in a two-dimensional space, using the PLS-loadings, PLS-partial weights or PCA-loadings vectors obtained from PLS or PCA models.
PCA-scores
Plots objects (molecules) in a two-dimensional space, using the PCA-scores vectors obtained from the PCA model.
If the PCA model has ever been used to produce external predictions using the command Utilities>>PCA-predictions, the dialog will show a control asking if we want to show the result of the predictions in the plot. If this control is selected, the plot will contains some additional points representing the objects in the external set, usually colored in red, at the possitions corresponding to their predicted scores.
Plot>>>2D plot>>>X val distribution
Plot>>>2D plot>>>X sd distribution
These two commands start a histogram application. Follow this link for a full description of their options and commands.
X val distribution
This histogram represents the values taken by any variable in the data file included in a X block.
X sd distribution
This histogram represent the standard deviation of any variable in the data file included in a X block.