TSCI.pngGraph Point Format (2D)

8052.png The Modify Point Format dialog for a graph incorporates a copy of the current graph. All changes made in any of the point format options are immediately reflected in the graph. The entry fields serve either the Y or Y2 axis, depending on whether Y Axis or Y2 Axis is selected. The Y2 options will not be available if the graph contains a single plot. This option sets only the principal data points of each Y axis. Reference data sets, if present, use a preset format.

If there are multiple graphs present, the point settings will apply to all of the graphs.


A graph's data point size may be set from zero to ten. Size 0 points are always pixels (non-bar shapes) or lines (bar shapes), regardless of the device. If the graph specifies a bar shape, this option affects the width of the bar.


A graph's points may be set as square, circle, diamond, triangle down, triangle up, or plus symbols. You may also choose a bar graph where the bar's base is zero, or the Ymin or Ymax of the plot.


You can choose to leave the points unfilled or fill them with one of the four point colors assigned to color schemes within the product.

For non-linear optimization graphs, you can choose Color by Residuals. Here the points are colored based upon the absolute value of the number of standard errors from the Y-predicted value. The default color schemes employ a coloring based upon ascending color wavelength. Points less than 1 standard error are shown in blue, those between 1 and 2 standard errors are green, those between 2 and 3 standard errors are yellow, and those beyond three standard errors are red. Points that are beyond 2 standard errors may represent outliers which may be adversely impacting the overall fit.


To make the points more visible, especially in contour plots, you can set the outline of a point to black or to one of the four point colors assigned to color schemes within the product.


You can elect to have inactive points filled or unfilled.

Points can be hidden entirely.

For non-contour 2D plots, a graph's points can be connected with lines, with a constrained cubic spline that defends against the overshoot typical of splines, or with a cubic non-uniform rational B-spline (NURBS) that offers a fixed smoothing.