
<strong>Importing and Viewing Data</strong>
<p>
To add a model (data product) to the display, use the 
Model&#8594;Load Model from File menu to 
select a file. You may also use a drag-and-drop gesture from your file system
(source data will be copied into application).
<p>
The viewer assumes that input files with the extension .txt are 
space-delimited text files (compatible with R).  Files with the
extension .csv are assumed to be comma-separated-value files (text files
with comma delimiters). The fields in each line should be given as 
x, y, and z coordinates. Additional columns are ignored. Vertex ID is 
assigned by line number.
<p>
The viewer can read Lidar directly from both LAS and LAZ files.  
<p>
Once data is rendered, you may use a click-and-drag gesture to pan the
image or a mouse wheel to zoom in and out.  Information and coordinates
extracted from the data source will be presented in the "readout" area
in the lower-right corner of the display.
<p>
The viewer attempts to manage the amount of clutter on the display
as well as rendering times but selecting subsets of the data when
it is appropriate to do so. As you zoom in, more and more data becomes available.
<p>
<strong>Display Options and Point Thinning</strong><br>
You may configure the display options by accessing the View menu.
Most of the options are self-explanatory, but the point-thinning options
require special attention. When you are dealing with a very large
data set, it is a good idea to use the point-thinning options enabled
so that the application will manage clutter and limit rendering times.
Wireframe rendering is particularly susceptible to slow rendering if
thinning is disabled.
<p>
On the other hand, enabling the full-resolution option for raster
images often improves their quality. This is particularly true when using the
hillshade option. However, activating this option requires the construction
of a full-resolution TIN and may lead to substantial memory use. In general,
Tinfour requires 124 bytes per sample (vertex) to build a TIN.  So a TIN
containing 1 million points requires roughly 124 megabytes. You should 
consider the size of your sample data and available Java memory when
using the full-resolution option. 
<p>
<Strong>Adding Constraints</strong><br>
Once a model is loaded, you may load constraints from a Shapefile or
text file by accessing the Model&#8594;Load Constraints from File menu
item. The viewer assumes that the constraints are in the same coordinate system
and cover the same area as the source model. If the data does not
conform to this requirement, the results are undefined.
<p>
When presenting the data with a large reduction factor,
the collection of points in the constraints may be reduced using
a geometry simplification (point reduction) technique
called Visvalingham's algorithm.

