Lineplots

A lineplot graphs a variable in case order and connects the successive points with lines. It is like a scatterplot of a variable against another that counts from 1 to n.

To make a lineplot, select the variable to plot and choose Plot > Lineplot.

Lineplots are often used for data recorded over time. However, it is a good idea to make a lineplot of any variable just to check for unexpected trends related to the sequence order in which the data are recorded. This is especially true if the cases are recorded in the order in which they were collected.

Lineplots space the datapoints evenly across the x-axis. If your data were measured in order but at unequal intervals, you may get a more appropriate picture of your data by adding lines to a scatterplot of the data against an x-variable that specifies the correct spacing.

The Multiple Lineplot command plots several variables, against a common sequence axis. Data Desk plots each line in a different color. Multiple lineplots show how several sequences move together. They do require, however, that all of the sequences be measured on the same scale or the y-axis of the plot is meaningless. Thus, for example, a multiple line plot of the price of Apple Stock (measured in dollars) and the number of Macintosh apples sold annually (in millions of bushels) might show related trends but would arbitrarily equate 100,000,000 bushels to $100.00. It is important to keep in mind that this equivalence is an artifact of the display and has no meaning for the data.

Multiple line plots must be used (if at all) with great care. They are most trustworthy when the sequences plotted are directly comparable, measured in the same units, and of the same general magnitude and range.

Multiple lineplots create a new relation holding three intermediate variables needed to construct the plot. The first of these appends the selected variables together. The other two provide the x-axis and group names. Because Multiple lineplots work in this separate relation they do not link to the relation of the original data and cannot work from the Ribbon. Initially, the relation is hidden. To make it visible, choose the Locate variable name command from the HyperView menu attached to either the x-axis or y-axis.