Indirect addressing eliminates the need for data transfer 
each time new data are required. This is accomplished by 
dimensioning the array which holds the band information in 
two dimensions - CONTOUR (n,2), where n is the maximum num 
ber of grid cells in one row. Each column represents one 
band of data and is addressed via the variables FIRST and 
SECOND. These variables contain either the value one or two 
depending upon which is the current working band and which 
is the neighbor band. Thus when new data are required the 
addresses in FIRST and SECOND are swapped and the new 
information is read into the array CONTOUR (n,SECOND). 
Utilization of these techniques provides for the rapid 
display of digital information in DEMGS. The approaches and 
techniques used in the different software 'package' used in 
DEMGS follows. 
Contour Band Ilap (DC3M) 
The contour band map (or choropleth map of elevation), con 
sists of two processes. The first is a classification of 
the data elements into the appropriate class interval, and 
the second is the display of the resultant information. In 
this type of display all elevations which are in a given 
class are drawn with the same color. 
In order to classify the data, the method of equal intervals 
was selected for maximum efficiency, since the algorithm for 
this process eliminates the need for any comparisions. The 
only information required was the elevation range, minimum 
elevation, and the number of class intervals. The first two 
elements can be derived from the header information in the 
H-file and the last is set to a maximum of sixteen (which is 
the number of colors available) but can be reset by the user 
to a lower value. The elevation range is calculated from 
the maximum and minimum elevations in the H-file. In order 
to classify the data an increment value is necessary as 
well, which is calculated as follows: 
INCREMENT = (MAXE - MINE) / NCLASS (1) 
Where, 
MAXE is the maximum elevation of the DEM, 
MINE is the minimum elevation, and 
NCLASS is the number of class intervals. 
Given this, the elevation class of a pixel is: 
CLASS = (ELEV - MINE) / INCREMENT (2) 
Where, ELEV is the current working elevation, and CLASS is 
the integer (truncated) value of the operation. 
After the classification process, which is accomplished 
during data transfer the information must be plotted. This 
is done in two stages. The first is to defer the actual 
drawing command until the classification changes by using a 
run-time run-length computation; this approach was taken to 
speed up the display since each DRAW command in the DEMGS 
hardware environment generates a CPU interrupt, thus by only 
drawing lines instead of points, the drawing time is min 
imized. The second stage is to use the class value as the