ano_obs_select

Select a given observation/set of observations from a UDB dataset.

%ano_obs_select(geo, time, idsn, odsn, var=, vartype=, where=, flag=, distinct=, ilib=, olib=);

Arguments

• geo : list of (blank separated) countries ISO-codes; note that when geo=_ALL_, all countries will be processed, i.e. no condition on ISO-code will be imposed;
• time : year of interest; note that when time=_ANY_, all will be processed, i.e. no condition on time will be imposed;
• idsn : (option) input dataset;
• odsn : (option) output dataset;
• var : (option) list of fields/variables of idsn upon which the extraction is performed; default: var is empty (or var=_ALL_) and all variables are selected;
• vartype : (option) type of variables passed, either personal (P), household (H) register (R) or (D); if empty (not recommended), the first letter of the first variable in var will be used in place of vartype;
• where : (option) expression used to refine the selection (WHERE option); should be passed with %str or %quote; default: empty;
• flag : (option) name of an additional flag variable (set to 1) added to the output dataset; default: ignored;
• distinct : (option) boolean flag (yes/no) set to use the DISTINCT option of the PROC SQL selection procedure;
• ilib : (option) input library; default: ilib=WORK;
• olib : (option) output library; default: olib=WORK.

Returns

Example

Imagine one needs to create a table with UK households containing more than 10 members (HHsize>=10), e.g.:

PROC SQL:
    CREATE TABLE work.removeH AS 
    SELECT DISTINCT
        HB020, HB030,
        (1) as remove
    FROM pdb.udbh 
    WHERE HB020 = 'UK' AND HHSIZE >= 10;
quit;

The procedure above can be ran equivalently using the following command:

%ano_obs_select(UK, _ANY_, udbh, removeH, var=HB020 HB030, vartype=H, 
    where=%quote(HHSIZE>=10), flag=remove, ilib=pdb);

See also

%obs_select, %ds_select.