Function to solve the non-linear optimization problem used
within ptable().
pt_optim_entropy(
optim = optim,
mono = mono,
v = v,
variance = variance,
lb = p_lb,
ub = p_ub,
ndigits
)The return value contains a list with two elements:
result"optimal value of the controls
iter" number of iterations that were executed
optimization parameter (1=default, 2-4=further test implementations)
(logical) monotony parameter
(integer) vector with perturbation values (i.e. deviations to the original frequency)
(numeric) variance parameter
(integer) vector with lower bounds of the controls
(integer) vector with upper bounds of the controls
(integer) number of digits
Tobias Enderle, Sarah Giessing, Jonas Peter
The main parameter is `optim`: In `optim=1 to 3` the variance is stated as inequality constraint and in `optim=4` the variance condition is stated as equality constraint.
Giessing, S. (2016), 'Computational Issues in the Design of Transition Probabilities and Disclosure Risk Estimation for Additive Noise'. In: Domingo-Ferrer, J. and Pejic-Bach, M. (Eds.), Privacy in Statistical Databases, pp. 237-251, Springer International Publishing, LNCS, vol. 9867.
Fraser, B. and Wooton, J.: A proposed method for confidentialising tabular output to protect against differencing. In: Monographs of Official Statistics. Work session on Statistical Data Confidentiality, Eurostat-Office for Official Publications of the European Communities, Luxembourg, 2006, pp. 299-302