Home / Column Articles/ October 2002 Column Article

October 2002 Column Article

Graphics Server .NET - Preview II

The Graphics Server wish list and developer support have seen numerous requests for a feature that allows you to overlay different chart types within the same X/Y axes extents. With Graphics Server .NET, this request has been addressed, so that most chart types can be combined/overlaid on the same graph. To change chart types, you simply point to the series (each set of data can be in its own series) and assign the chart type you want for that series. The example image shows multiple chart types combined into one graph, i.e. within the same X/Y axes extents.

Notice how this first image has combined a line, bar, area and scatter graph. The bar series has some transparency turned on so you can see the area series behind it. Also, note how the legend shows different icons for each series to visually distinguish one series from the next.

Another example that is often used in stock chart reporting is an Open-High-Low-Close chart type combined with a Volume chart type. The Open-High-Low-Close series shows the range of a stock's price for a given day (the light blue series in the above example), and the Volume chart type shows the amount of stock traded for the same day (the light green series). Note that a Limit Line is used, (the a dashed line pattern) to separate the Open-High-Low-Close series from the Volume series. Notice how the legend in the second example uses colors to differentiate each series rather than the usage of symbols like in the first graph. Of course, differentiation combing both the use of colors and symbols is possible.

Trend lines are another feature that has been improved in Graphics Server .NET. All the trend lines from Graphics Server 6.0 are again available, but the enhanced feature in .NET is the ability to add as many trend lines as desired to the same data series. The first trend line example image shows three data series that each have their own max/min and mean lines.

Notice that each trend line can have its own color (currently the same color as the series the trend is associated with), along with its own line pattern (dash line pattern or dash-dot line pattern).

The last example shows the power of the trend lines, showcasing a single series with six variable order polynomial curves associated with that series.

One can see that as the polynomial order of the curve increases, the better the fit to the series data. A polynomial order of 1 doesn't fit the data too well (see the red trend line), while a polynomial order of 11 fits the data exactly (see the blue trend line - the same color curve as the series data color). This example only shows the variable order polynomial trend, but you could just as well add Logarithmic trend lines, Inverse trend lines, and other trends either on top of the current trend lines, or in place of the current trend lines.