anderej (05/26/82)
I had hopped that a statistics group would be arround that I could post this to, but..... I am trying to bring an S system up on an 11/70 running V7 UNIX(TM). Everything was going well (sort of) until I got one of the (hopefully soon to be) users of the package to take it out for a spin. The problem occured in the regress function, and may occur elsewhere. When the users environment is sufficiently small (ie just PATH, HOME, PS1 etc) the arguments to ientry in the file $SHOME/s/source/u/wtout.r get written over. Large environments I believe only mask the problem. I have been trying to solve this for some time and would appreciate any and all assistance. Thanks in advance, Eric J. Anderson ...!harpo!stolaf!anderej ...!ihnss!ihps3!stolaf!anderej