jj (06/07/82)
My most pressing interest in a caving newsgroup is for the purpose of discussion loop cloture in surveys by whatever methods. I have worked out and tested a least squares method of loopp closure that uses covarient weighting in all three dimensions. It has been tested only on contrived data due to the fact that I have neither the time nor the expertise to generate the I/O routines necessary to use the program as a surveying aid. My abilities(if any) lie in the mathmatical end of the problem. For those of you who are wondering what covarient weighting does, I will attempt to explain. A survey shot done by brunton (or suunto) , clinometer, and tape, does not have equal variances of error in x,y,and z dimensions. In addition, the axis of the error do not, as a rule, lie on the x,y, and z axis. The covarient method of weighting uses the theoretical (sigh, that word) weighting for the shot when doing the least squares fit. Most least squares fits do each axis separately, and lose a small bit of the information. As far as I can tell, the more comples covarient weighting provides a better result in caves where an absolute measurement is required, such as in surveys that hope to make connections, etc. Any other mapping people out there? Aside from that interest, a general summary of events, discoveries, and so on would also be a reasonable thing to include in this newsgroup. /