Housing units are known as multidimensional goods, yet the approach of identifying this multiple characters of properties are seldom clarified. This study allocates a perspective of identifying the many dimensions of housing unit from spatial point of view. The concept of proximity or nearness in particular is essential in many forms of human reasoning. Distance is then one of the factors that determine proximity, but not the measure of proximity itself. While the proximity measure is mostly a conceptual formation, it is often inversely proportional to a distance measure. This study identifies appropriate measures of distance using Nearest Neighbourhood Analysis (NNA) to generate spatial attributes. Kernel Density Estimators (KDE) is later used to illustrate patterns of spatial clusters for housing units’ distances patterns. The analyses show distance matrices as appropriate benefactor to derive appropriate spatial attributes, while the diversity of Kernel Density patterns shows different spatial perspectives of housing units. The implications of monocentric and polycentric patterns determine contributions of distance matrices in multidimensional aspect of a housing unit.