[From Bill Powers (950913.1415 MDT)]
Slight error in the program I posted earlier. The first line of code in
the dist function has an extra term that got put in by a cosmic ray
(determined after extensive research). The function should read
function dist: real;
begin
x1 := 10000.0*(random - 0.5);
x2 := x2 + slow * (x1 - x2);
x3 := x3 + slow * (x2 - x3);
dist := x3;
end;
For some reason taking it out made the RMS errors larger than before for
smaller values of the slowing factor (possibly due to rounding errors in
adjusting the table for zero average and 1000 max value):
Slowing = 0.01 RMS ERROR = 5.5
Slowing = 0.02 RMS ERROR = 11.0
Slowing = 0.03 RMS ERROR = 15.9
Slowing = 0.04 RMS ERROR = 17.9
Slowing = 0.05 RMS ERROR = 23.2
Slowing = 0.06 RMS ERROR = 25.7
Slowing = 0.07 RMS ERROR = 31.8
Slowing = 0.08 RMS ERROR = 36.2
Slowing = 0.09 RMS ERROR = 37.8
Slowing = 0.10 RMS ERROR = 42.9
Slowing = 0.11 RMS ERROR = 48.7
Slowing = 0.12 RMS ERROR = 52.1
Slowing = 0.13 RMS ERROR = 59.6
Slowing = 0.14 RMS ERROR = 58.1
Slowing = 0.15 RMS ERROR = 66.1
Slowing = 0.16 RMS ERROR = 70.9
Slowing = 0.17 RMS ERROR = 66.9
Slowing = 0.18 RMS ERROR = 80.7
Slowing = 0.19 RMS ERROR = 82.9
Slowing = 0.20 RMS ERROR = 86.4
ยทยทยท
-----------------------------------------------------------------------
Best,
Bill P.