| Abscissa (x-axis) |
| |
You may select from linear, logarithmic, and date/time axes. Note, that logarithmic axes are only possible, if the entire axis is positive and does not include zero. |
| Ordinate (y-axis) |
| |
You may select from linear, logarithmic, and date/time axes. Note, that logarithmic axes are only possible, if the entire axis is positive and does not include zero. |
| Type of Plot |
| Col/Idx |
plot the data of a single column of the data matrix against the row index. |
| Row/idx |
plot the data of a single row of the data matrix against the column index. |
| Col/Col |
plot the values of two columns against each other. |
| Row/Row |
plot the values of two rows against each other. |
| Histogram |
display a histogram of selected parts of the data (see below for details) |
| Cumul.Frequ. |
calculate and plot the cumulative frequency distribution of selected parts of the data matrix. |
| Norm.Prob.Plt. |
creates a normal probability plot. |
| Pareto Chart |
creates a Pareto plot. |
| Labeling of Axes |
| Persistent |
The labels of the axes are fixed. |
| Numeric |
The column or row numbers are displayed. |
| Names |
The variable or object names are displayed. |
| Mode of Display |
| Lines |
line plot, connecting all data points in the order of their occurence in the data matrix |
| Points |
point plot, any data item is displayed as a point whose type can be setup by clicking the button besides this option. |
| Spectrum |
display data as a line spectrum. The data points are shown as spectral lines which are based on the horizontal axis (y=0). This mode is not available for all types of charts. |
| Classified Line |
show data as a line drawing, connecting all data points of the same class. This mode can be used to create distinct overlaid graphs. |
| Attributes |
| Any data points may be displayed using one of the following attributes. In addition, the data can be colored according to the class number. |
| None |
displays the data as is, without any additional attributes. |
| Data Points |
the locations of the data points are indicated by plotting a symbol for each data point on top of the base graph. You can select the symbol by clicking the button right to this option. |
| Numeric |
the data points are displayed by using numeric values which indicate either the row or the columns index of the particular point. |
| Class Symbols |
the class numbers are indicated by different symbols. |
| Class Numbers |
the class numbers are displayed as numeric values (range 0 to 127) |
| Class Colors |
if this checkbox is checked the data points are additionally marked by colors which are derived from the corresponding class numbers. The class colors override the default data color. Note that the class colors can be assigned on an individual basis by using the command Setup/Colors/Class Colors. |
| Colors |
| The data plots of DataLab are based on four different colors, which can be set up individually for each plot: |
| Data |
this color is used to display the data unless the attributes are set to 'Class Colors' |
| Fill |
used for filled rectangles (as in histograms) |
| Scale |
color of the scales |
| BkGnd |
background color |
| Grid |
grid color. Be sure to select a color different from the background color, otherwise the grid would not be visible |
| Options |
| Display Grid Lines |
When checked grid lines are displayed. The distance between lines follows the tick marks of the scales. |
| Do not Draw Superimposed Symbols |
If you intend to display large datasets (more than 10000 data points) this may result in a notable reduction of the drawing speed. Ticking off this option prevents data points being drawn on top of each other (in the immediate neighborhood of +/- 2 pixels). This increases the drawing speed considerably (starting with about 100000 data points you'll want to activate this option).
|
| Zoom Mode |
The Zoom Mode determines whether DataLab automatically adjusts the zoom range to fit all data to be displayed. If Zoom Mode is set to fixed the user can specify the range of the displayed data by entering the borders of the intended range. This range will be retained until changed by the user. Clicking the "zoom to scale" button  adjusts the zoom range such that all data points are visible and both scales start and end with a tick mark.
If Zoom Mode is set to automatic, the zoom range is automatically adjusted whenever new data are displayed. |
| Data Range |
| This section is only visible if the plot type has been set either to "Histogram", "Cumul.Frequ.Dist.", or "Norm.Prob.Plot". You may select which parts of the data matrix are to be used for the calculation of the respective plots (column, row, marked data or entire matrix). In case "Column" or "Row" is selected, the column or row to be used can be selected as usual (see the description of the plot windows on switching the axes). For histograms you can additionally select whether to display the class information in the chart (option "Classified Data" |
| Histogram Parameters |
| This section is only visible if the plot type is set to "Histogram". You may select which parts of the data matrix are used for the calculation of the histogram, and the number of histogram bins, and the depth and the angle of the bars. Setting the latter two parameters to non-zero values creates 3D bars. If you tick off the "Show Normal Density" box a normal distribution having the mean and the standard deviation of the displayed data is drawn over the histogram.
There are three ways to specify the histogram classes:
- Automatic: If you select "automatic" the number of classes will be set in a way (depending on the type of the variable and the number of data) that the data structure is reflected by the histogram in the best possible way. The automatic mode is best suited for getting a quick overview. Special needs can be met using the other two options.
- Fixed Number of Bins: In this mode you can set the number of classes by moving the slider. The bin width and the data range are then calculated by DataLab automatically.
- Exact: This option allows to set the origin and the width of the classes precisely. The boundaries of the class bins are calculated such that they form a series of integer multiples of the class width, including the origin.
|
Features of DataLab 
Use the options key in any plot window