The data can be fitted to (10) by iteratively using a standard
least squares procedure. Care is required with the non-linear parameter
. The quality of the fit can be tested by computing 
defined as
where 
runs over all data points and 
is the error in
point . After fitting should be approximately equal to the
number of data points less the number of fitted parameters. Hence the
value of provides a measure of the quality of the fit. In
the results presented here the range of values of disorder round the
critical value was chosen such that conforms to this
condition. Then a large number of additional points was calculated
inside this range. An important side effect of this procedure is that
the apparently acceptable range of disorder around the fixed point gets
narrower
as the calculations become more accurate. It is therefore important to
test whether any apparent change in the fitted exponent is due to this
narrowing.
The values of the ideal and the fitted as well as the range
considered are shown in table 1. Using 
and the
widest range of disorder 
and 
for rectangular and Gaussian cases respectively.
