| Data Analysis, MATLAB® Version 7.6 (R2008a) Release Notes | |
New features and changes follow.
This release introduces two new interactive tools for data exploration:
Data brushing — For marking observations on graphs, allowing you to remove or save them to new variables
Data linking — For connecting graphs with data sources (workspace variables) to automatically and interactively update them
In addition, figure windows have a new banner, called the linking and brushing message bar. By default, when you plot data into a figure (i.e., add axes), an informational banner appears across top of the figure that looks like this.
Click the first two links to read about these new tools. Click the Play video link to open a nine-minute video tutorial about the tools in a browser window (the video also describes new GUI-building features.) To dismiss the banner, click the X. Once you do, the banner only reappears on subsequent plots if you select Show linking and brushing message bar in the MATLAB® Preferences Confirmation Dialogs panel.
Note The linking and brushing message bar can obscure a plot's title. Also, if you do not dismiss the message bar, it is visible in images of figures captured with getframe, but it does not print. |
Use data brushing when you want to isolate observations in a 2-D or 3-D graph for separate analysis, or to remove outliers or noisy data points. Data brushing can be applied to most graphs (some plot types do not support brushing). Data brushing is an exclusive, persistent mode. That is, when using it, you cannot use other figure tools, but the results of brushing data persist when you select a different tool or no tool.
Use data linking to make plots dynamically respond to changes in the variables they plot. Data linking applies to most graphs with identifiable data sources and operates at the figure level. Data linking is not modal and persists until you toggle it off or the connection between a plot and its data sources is broken.
The two tools work smoothly together and with the Variable Editor to visually highlight brushed observations and the data values they represent:
Brushing data-linked observations on a graph highlights them on other graphs that display them.
Brushing highlights values in the Variable Editor when a brushed variable is displayed there.
Using the Brush tool in the Variable Editor highlights values you brush that appear in linked plots.
Changing values of variables causes linked graphs displaying them to update with the changes.
Clearing variables disconnects them from all linked figures displaying graphs of them
You can modify variables from the command line, the Variable Editor, or with M-files. When used within functions, data linking operates in the function's workspace, not the base workspace. This is also the case when debugging.
All figure windows that contain axes now include a data brushing
tool (the Brush/Select data icon
) that lets you enter and exit brushing
mode and select a color with which to brush observations. The tool
draws selection rectangles (in 2-D plots) or prisms (in 3-D plots)
and permits you to select discontiguous regions and negate previously
brushed observations. Undo is also supported.
The Brush/Select data tool is a "split button" control with a brush icon on the left and a drop-down color palette on the right. When you depress the brush icon, you are in brushing mode; all data observations you select are highlighted with the current brush color. The figures below illustrate these operations.
|
|
If you leave data brushing mode to zoom, pan, or edit the plot, all brushed observations remain highlighted. You can then reenter brushing mode and pick up where you left off. Brush marks are not preserved when you save a figure and reopen it from the FIG-file, however.
Use the brush function to turn brushing on and off, and to select a color for brushing graphs. You can change brush colors on the fly with either the API or with the Brush tool.
All figure windows that contain axes now include a data linking
tool (the Linked Plots button
) to toggle linked mode
on and off (the default). When you toggle it on, an information bar
appears underneath the lowest toolbar on the figure, as shown below.
It displays what variables are linked to each series (data sources
for x-, y-, and z- data in the graphs).
On the left side of the information bar is a drop-down menu
that displays the symbolism and identifies
the data source for each series currently linked. On the right side
is an Edit button that opens the Data Source
Properties dialog box in which you can set display names and data
sources. Usually it is possible to unambiguously determine what data
sources a graph has, but sometimes you need to indicate what data
source to use, for example, when you plot a subrange of a data array.
The information bar explains that you need to do this as soon as you
turn on data linking; then, you can open the Data Source Properties
dialog box to identify your data source(s).
Use the linkdata function to turn data linking on or off for the current figure or for a figure for which you supply a handle.
| Mathematics, MATLAB® Version 7.6 (R2008a) | Programming, MATLAB® Version 7.6 (R2008a) | |
| © 1984-2008- The MathWorks, Inc. - Site Help - Patents - Trademarks - Privacy Policy - Preventing Piracy - RSS |
