n.a. (1977). Comput Appl Archaeol 1977.

Up to now, 'best' circles and ellipses for fitting to megalithic constructions have been derived by trial-and-error methods, a statistical method is now presented which uses coordinate point data from the site to produce a 'best fit' circle or ellipse. A second technique produces an alternative geometry for Thom's flattened circles and egg-shaped rings. On the ground these can be produced by a rope thrown round three pegs placed in an isosceles triangle. A computed catalogue of possible shapes resulting from such a method was produced as a comparison for real-site data, so that by iterative means an apparent best fit may be found for any given site. However, neither this method nor Thom's complex geometry necessarily reproduces the method used by the original builders.

Using data from an early post-Pleistocene open site (Liencres, Spain) nearest neighbour analysis was performed and graphic output obtained which allowed for objective measurement of spatially overlapping clusters of objects. The method was to compare the distributions of those objects with random, maximally dispersed and maximally aggregated theoretical distributions adjusted for density. Jaccard's coefficient, chi-squared and Pearson's contingency coefficient were used (1) to compare tool frequencies occurring in 'shared space', (2) to evaluate the (statistical) significance of point scatters and (3) to measure the strength of relationship between spatially co-occurring pairs of tool types. Au(abr)

Deduces that the monothetic method is capable of producing an infinite number of 'classifications'; but that by careful choice of weighting for presence as opposed to absence and of weights for individual attributes, the monothetic method may be made to produce results as acceptable as those from the polythetic method, with the added advantage that the specific attributes responsible for the discrimination between groups are identified at all stages of the process. Au

The results of canonical variates analysis applied to data from the best documented Mousterian assemblages from SW France support the division of French Mousterian industries into the four major groupings originally proposed by Bordes. CB

Presents a new method for the quantitative spatial analysis of occupation floors, or more generally of archaeological horizons. The method is directly applicable to 'precise' coordinate data, counts for quadrats or irregular areas, and to an intermediate form of data involving collection by a sub-grid which effectively yields 'approximate' coordinates (typically + 15cm). It is suggested that the second method is often ideal for archaeological applications where 'precise' coordinates may merely resolve post-depositional noise or juxtapositions which should be noted during excavation. It is argued that local density analysis is the only method capable of treating such data, and the data sets necessary and programs available are discussed. Au

The ancient monuments protection service of the Rhineland has installed a system for transforming archaeological information in oblique aerial photos to 1:5000 base maps. A comprehensive software-hardware plan was implemented and employs windowing, enhancement, pseudo-colour density slicing and merging. A simple map compression scheme which offers a reasonable compromise between bit rate and computing time has been developed. Au