Plot Options Tab
Style. Select either Point or Line for the plot style. Selecting Line activates the X Mode control. Selecting Point will use a plus symbol (+) as the default point shape. Once the graph is created, you can change the point shape and alter its size.
X Mode. For line plots, you must choose an X Mode to define the style of the line plot. Select Sort, Overlay, or As Read. See the topic Plot Node for more information. For Overlay or As Read, you should specify a maximum dataset size used to sample the first n records. Otherwise, the default 2,000 records will be used.
Automatic X range. Select to use the entire range of values in the data along this axis. Deselect to use an explicit subset of values based on your specified Min and Max values. Either enter values or use the arrows. Automatic ranges are selected by default to enable rapid graph building.
Automatic Y range. Select to use the entire range of values in the data along this axis. Deselect to use an explicit subset of values based on your specified Min and Max values. Either enter values or use the arrows. Automatic ranges are selected by default to enable rapid graph building.
Automatic Z range. Only when a 3-D graph is specified on the Plot tab. Select to use the entire range of values in the data along this axis. Deselect to use an explicit subset of values based on your specified Min and Max values. Either enter values or use the arrows. Automatic ranges are selected by default to enable rapid graph building.
Jitter. Also known as agitation, jitter is useful for point plots of a dataset in which many values are repeated. In order to see a clearer distribution of values, you can use jitter to distribute the points randomly around the actual value.
Note to users of earlier versions of IBM® SPSS® Modeler : The jitter value used in a plot uses a different metric in this release of IBM SPSS Modeler. In earlier versions, the value was an actual number, but it is now a proportion of the frame size. This means that agitation values in old streams are likely to be too large. For this release, any nonzero agitation values will be converted to the value 0.2.
Maximum number of records to plot. Specify a method for plotting large datasets. You can specify a maximum dataset size or use the default 2,000 records. Performance is enhanced for large datasets when you select the Bin or Sample options. Alternatively, you can choose to plot all data points by selecting Use all data, but you should note that this may dramatically decrease the performance of the software.
Note: When X Mode is set to Overlay or As Read, these options are disabled and only the first n records are used.
- Bin. Select to enable binning when the dataset contains more than the specified number of records. Binning divides the graph into fine grids before actually plotting and counts the number of points that would appear in each of the grid cells. In the final graph, one point is plotted per cell at the bin centroid (average of all point locations in the bin). The size of the plotted symbols indicates the number of points in that region (unless you have used size as an overlay). Using the centroid and size to represent the number of points makes the binned plot a superior way to represent large datasets, because it prevents overplotting in dense regions (undifferentiated masses of color) and reduces symbol artifacts (artificial patterns of density). Symbol artifacts occur when certain symbols (particularly the plus symbol [+]) collide in a way that produces dense areas not present in the raw data.
- Sample. Select to randomly sample the data to the number of records entered in the text field. The default is 2,000.