Attenuation matrix in robust, free adjustment

Equivalent weight matrix Px plays a major role in robust, free adjustment. It is contained in the optimization criterion P(dx) = dtxPxdx where dx is an increment vector to approximate coordinates of all network points. Assuming, that Px = px T(dx ), where Px is a priori weight matrix, the paper presents the way how to calculate an attenuation matrix T(dx)( dx is a standardized increment vector). Special attention is paid to the way of increment standardization and to computation of an increment variance matrix.