Why are 1-D arrays used in Analytic Surface?

[wobster]

Hi,

I've been looking at the Analytic Surface demo (shown below) and was curious what the motivation was to use 1D arrays for what is essentially a more natural 2D set of data. I imagine it had to do with passing the buffers around in an optimal manner, but I'd be interested to know the full architectural motivation and optimization trade-offs for this for my own edification

Thanks!

Rob


package gov.nasa.worldwind.examples.analytics;
class AnalyticSurfaceDemo ... {

public static BufferWrapper randomGridValues(int width, int height, double min, double max, int numIterations,
double smoothness, BufferFactory factory)
{
int numValues = width * height;
double[] values = new double[numValues];

for (int i = 0; i < numIterations; i++)
{
double offset = 1d - (i / (double) numIterations);

int x1 = (int) Math.round(Math.random() * (width - 1));
int x2 = (int) Math.round(Math.random() * (width - 1));
int y1 = (int) Math.round(Math.random() * (height - 1));
int y2 = (int) Math.round(Math.random() * (height - 1));
int dx1 = x2 - x1;
int dy1 = y2 - y1;

for (int y = 0; y < height; y++)
{
int dy2 = y - y1;
for (int x = 0; x < width; x++)
{
int dx2 = x - x1;

if ((dx2 * dy1 - dx1 * dy2) >= 0)
values[x + y * width] += offset;
}
}
}

smoothValues(width, height, values, smoothness);
scaleValues(values, numValues, min, max);
BufferWrapper buffer = factory.newBuffer(numValues);
buffer.putDouble(0, values, 0, numValues);

return buffer;
}