We introduce a method for guiding interactive exploration of high-dimensional
data. The method is based on nine characterizations of the 2D distributions
of orthogonal pairwise projections on a set of points in multidimensional
Euclidean space. These characterizations include measures such as, density,
skewness, shape, outliers, and texture. Using with these measures, we can
quickly generate a comprehensive summary of the 2D relations of variables in
a large dataset with more than a hundred dimensions.