iiplot

iiplot#

Purpose

Refresh all the drawing axes of the iiplot interface.

Syntax

iiplot

Description

iiplot is used to scan through multiple sets of 1D (function of time or frequency) and 2D responses (functions of two variables) as discussed in Type. Section 2.1 gives an introduction to the use of iiplot and the companion function iicom.

The data is stored in a Stack using one of the accepted curve formats.
iicom CurveInit is the base command to add curves in the stack. You can also create a new iiplot axis using a curve data structure Curve (generated by fe_curve for example), simply calling iiplot(Curve).
Each iiplot axis (see iicom Sub, ) can display some or all data sets in their stack. The selection of what is displayed is obtained using the iicom IIx commands or the Variables menu.
iiplot with no arguments refreshes all the drawing axes.
Plot Type supported by iiplot are described below. The plot type can be selected using the PlotType menu of the toolbar or through iicom Show commands.
Selected channels (columns of the data sets) are shown for all plots. The iicom commands +, -, ch and the associated keys and toolbar buttons can be used to change selected channels.
Pole lines for the indication of pole frequencies, or other lines to be shown (harmonics, thresholds, ...), are available for many plots. In general the information for these lines is stored as a Curve.ID field. The IIplot:PoleLine menu can be used to change how these lines appear.
For identification (see idcom) ci.Stack{'IdMain'} pole lines are shown in white/black.
ci.Stack{'IdAlt'} pole lines in red.

ci : handle#

ci=iiplot returns a SDT handle to the current iiplot figure (2nd optional output argument is XF, a pointer to the curve stack, see section 2.1.2). You can create more than one iiplot figure with ci=iiplot(FigHandle).

PlotInfo#

Curves to be displayed can contain a C1.PlotInfo cell array of stringTag,data. An alternate form using matrix where the first column gives tag and the second the data is accepted if that matrix has at least two rows.

LineProp specifies properties to be used as properties for lines. For example C1=sdsetprop(C1,'PlotInfo','LineProp',{'LineWidth',2}). This is checked at each display.
sub, show, scale commands to be executed when initializing a display tab with iicom Sub.
ua.PostFcn commands executed at the end of a refresh. This gives the user a chance to introduce modifications to the result of iiplot.
ua.TickFcn commands executed whenever a mouse zoom is done, see TickFcn.
ua.axProp is a cell array containing properties to be applied with an comgui objSet command every time the plot is refreshed.

OsDic is a string or a cell array of OsDic entries defining figure naming, colorbars (see also fecom ColorBar), line styles, ... done once at the CurveInit call.

 C1=d_signal('Resp2D');
 % See sdtroot('InitOsDic') or sdtweb('osdic')
 C1=sdsetprop(C1,'PlotInfo','OsDic',{'CbTRRNoLab','FnI'});
 iicom('curveinit','2D',C1);

LDimPos specifies the dimension used to generate the label on the response axis (y for f(x), z for f(x,y)).
{'ch',index}, {'ch','all'}, {'ch',{'Xlab',index}} can be used to initialize channels.

The ii_plp('PlotInfo',C1) command provides default values for classical configurations.

Type#

2D (f(x)) plots are associated with the following buttons and iicom Show commands Abs (absolute value), Pha phase, Phu unwrapped phase, Rea real part, Ima imaginary part, R&I real and imaginary, Nyq Nyquist.
3D (f(x,y)) plots are image, mesh, contour and surface. For this plots ua.XDimPos should give the positions of dimensions associated with the x and y variations. Proper .PlotInfo can be generated with ii_plp('PlotInfo2D -type "contour"',C1).

DimPos and channel#

When displaying multi-dimensional curves as 2D plots f(x), the abscissa x is taken to be the first dimension declared in the C1.DimPos field (with a default at 1).

When displaying as 3D (f(x,y)) plots, the x,y are taken to be the first two dimensions declared in the C1.DimPos field (with a default at 1,2). You can then flip the positions in the plot axis by setting ci.ua.XDimPos=[2 1].

Channels are indices for remaining dimensions.

The y (z for 3D) axis label is built using the C1.DimPos(2) dimension unless the curve contains a LDimPos entry.

TypeIDcom#

Specialized plots for idcom are

Local Nyquist plots (initialized by show lny) show a comparison of Test (measured FRFs) and IdFrf (identified model) in a reduced frequency band

(10.29)

near the currently selected pole (the current pole is selected by clicking on a pole line in another plot axis). Local Nyquist plots allow a local evaluation of the quality of the fit. The error and quality plots give a summary of the same information for all the frequency response functions and all poles.
Multivariate Mode Indicator Function (initialized by show mmi), forces associated to the MMIF (initialized by show fmi), Alternate Mode Indicator Function (show ami), and Channel Sum (show sum) are four ways to combine all the FRFs or a set to get an indication of where its poles are located.
These indicators are discussed in the ii_mmif Reference section. They are automatically computed by iiplot based on data in the 'Test' set.
Pole locations in the complex plane (initialized by show pol).
Poles shown as damping vs. frequency are initialized by show fre.
Position of residues in the complex plane are initialized by show cre. This plot can be used to visualize the phase scatter of identified residues.
Value of real residue for each measured channel are initialized by show rre.
Error Local Nyquist error (initialized by show err). For the current pole, error plots select frequency points in the range [ω_j(1−ζ_j) ω_j(1+ζ_j)]. For each channel (FRF column), the normalized error (RMS response of ci.Stack{'Test'}.xf - ci.Stack{'IdMain'}.xf divided by RMS response of ci.Stack{'Test'}) is shown as a dashed line with + markers and a normalized response level (RMS response of ci.Stack{'Test'}) as a dashed line with x markers.
Normalized errors should be below 0.1 unless the response is small. You can display the error using the nominal sensor sort with ci.Stack{'IdError'}.sort=0 and with increasing error using sort=1.
Quality Mode quality plot (initialized by show qua), gives a mean of the local Nyquist plot. The dashed lines with + and x markers give a standard and amplitude weighted mean of the normalized error. The dotted line gives an indication of the mean response level (to see if a mode is well excited in the FRFs). Normalized errors should be below 0.1 unless the response is small.