ano_var_btcode

%ano_var_btcode(geo, time, db, var, wgt, db_d=, vartype=, coding=T, nobs=10, lib=WORK);

Examples

In the following, we consider the example of anonymisation applied in practice for geo=SI. The input datasets are the common UDB files whose extension states whether they contain R, P, H or D variables.

1. Variable top-coding: say for the highest 10 original values of the variable PY031G, you want to replace the original values with their weighted average. Run:

%ano_var_btcode(SI, UDB_P, PY031G, PB040, coding=TOP, vartype=P, nobs=10);

1. Combined variables top-coding: say you want to replace the values of PY030G variable for observations that are highest for either PY030G or the related variable PY031G, doing the following:
   • selecting the 10 IDs with the highest values of variable PY030G;
   • selecting the 10 IDs with the highest values of related variable PY031G;
   • considering the union of selected IDs (at least 10 observations, not more than 20),

then replacing the original values for observations in the union above with weighted average, you can then run the following:

%ano_var_btcode(SI, UDB_P, PY030G, PB040, coding=TOP, relvar=PY031G, nobs=10, vartype=P);

1. Gross/net variables top-coding: say you want to simultaneously replace the highest original values of gross/net variables HY040G/HY040N, doing the following:
   • selecting the 10 IDs with the highest original values of gross variable HY040G;
   • selecting the 10 IDs with the highest original value of net variable HY040N;
   • considering the union of selected IDs (at least 10 observations, not more than 20);

then replacing the original values of both gross/net variables for observations in the union above with their respective weighted averages, you can then run the following:

%ano_var_btcode(SI, UDB_H, HY040G HY040N, DB090, coding=TOP, db_d=UDB_D, nobs=10, vartype=H);