|DataLab is a compact statistics package aimed at exploratory data analysis. Please visit the DataLab Web site for more information....|
|Home Features of DataLab Mathematical/Statistical Analysis Correlation Scatter Plots|
|See also: Guided Tour: Histograms, Scatter Matrix
Scatter plots offer a simple and efficient basis to quickly recognize relations among variables. DataLab allows to display and evaluate up to five variables simultaneously. In order to open the scatter plot window you have to click the command Math/Bivariate Statistics/Scatter Plots (or the toolbar button ):
The scatter plot window at the left shows the matrix of the scatter plots of up to 5x5 variables. The diagonal of the matrix of scatter plots contains stacked dot plots of the corresponding variables in form of line spectra.
In many cases the data matrix contains significantly more than 5 variables. In this case the viewport of the scatter plot matrix can be panned with the scrollbars right to and below the plot matrix. If you prefer not to show all variables in the scatter plot window, you can (de)select interesting variables in the list of variables. If you select less than 5 variables the scatter plots will grow in size accordingly.
In order to display the regression line between each pair of variables you have to tick off the "Show Reg. Line" checkbox. The parameters of the regression line can be obtained by double-clicking a particular scatter plot. This displays a protocol window showing the most important uni- and bivariate parameters, including the slope and the intercept of the regression line. The button "Cross Corr. Table" shows the cross correlation table of all variables. If data belong to a time series it may be helpful to connect the data points by a line (checkbox "Connect Dots")
One question which is raised quite often is how the points in different scatter plots are related to each other. In order to provide support for this question DataLab allows to ink individual objects. The inked objects are displayed in parallel in all scatter plots as colored squares and can be marked by clicking the buttons or .
The inking of the data can be achieved as follows: simply click into one of the scatter plots and hold the left mouse button pressed while dragging it - this will show a rectangular region which follows the mouse movement. All objects within the rectangle will be inked when the mouse button is released. Normally, any existing inked objects will be reset when a new rectangle is defined. If you want to add additional objects you have to press the "Ctrl" key when dragging the mouse. The inking can be reset by clicking the button .