north, south, east, west not clearly defined

[mankoff]

Given a small sample input file like:

0,0,10
10,10,10

and running p2g with a very large search radius and resolution, one would expect a single output point with a min, mean, and max of 10. This works. But what is the definition of north, south, east, and west in this case?

The output file says they are all 10, but I would expect them to be the cell edges [0,10] or cell center 5.

[mankoff]

Related to #7 and better described there, although that bug report focuses on the Z value, whereas this focuses on the N/S/E/W values.

[gadomski]

I'm not sure if the cell edges should be [0, 10] in this case — if you're specifying a very large resolution (you used 1e4 in other issues, e.g. #5), shouldn't the cell edges be controlled by that resolution?

However, I'm going to keep this issue open, since the definition of proper behavior for your example is poorly defined. I agree that it probably should be something sane (centered on the centroid of the points? using the min point as the lower left?), and right now it's very arbitrary how this is handled.