The VAPOR Transfer Function Editors include histograms, that can be useful in understanding variable data. The histogramming capabilities include the following:
- Data can be histogrammed within any box extents in the scene by use of one of the transfer function editors.
- To histogram the data values in a 3D axis-aligned box, specify the box extents in the region panel (or use Region mode to specify the extents). Then use the Transfer function editor in the DVR panel, specifying the variable to be histogrammed as the DVR variable, and click the "Histo" button to view the histogram of the data values.
- To histogram a 2D variable, use the 2D box in the 2D panel to specify the extents to be histogrammed, set the 2D variable and use the transfer function in the 2D panel to see the histogram.
- To histogram a 3D variable in an arbitrarily rotated box, position the Probe box to control the box extents and orientation.
- The interval to be histogrammed can be specified using the "TF Domain Bounds" below the transfer function editor. Click the button "Fit Data" to set the histogram limits to include the full range of values of the variable.
- Note that the values in the histogram represent the distribution of variable values at the LOD and Refinement Level specified in the tab containing the Transfer Function editor.
- The histogram can be scaled in various ways: Click on the left mouse button in the editor window, and choose either "Boolean", "Linear", or "Log" scaling of the histogram. The boolean scaling is useful to identify where any data points are. The log scaling provides enables both large and small histogram values to be seen together.
- If the histogram shows only one positive value, it is likely that that value dominates all the values in the histogram. Less frequently occurring values can be seen by setting the histogram scaling to "Log" or "Boolean".
- The histogram height is normalized so that the maximum histogram count reaches to the top of the window. Hold the mouse over a point in the histogram to see the actual count associated with the value.
- The histogram always divides the histogrammed interval into 256 bins, evenly dividing the interval between the TF Domain Bounds. More accuracy can be obtained by restricting the TF domain bounds.
