Efficient evaluation of query expressions including grouping clauses

One or more non-transitory storage media storing a sequence of instructions which, when executed by one or more computing devices, cause performance of a method for aggregating a measure column vector according to a plurality of grouping column vectors, said measure column vector and each of said plurality of grouping column vectors being encoded column vectors, the method comprising:

• determining a plurality of actual grouping keys, each actual grouping key occurring in said plurality of grouping column vectors;
• generating a dense identifier mapping that maps said plurality of actual grouping keys to a plurality of dense identifiers, wherein said dense identifier mapping maps each actual grouping key of said plurality of actual grouping keys to a respective dense identifier of said plurality of dense identifiers;
• wherein each dense identifier of said plurality of dense identifiers corresponds to a respective workspace of a plurality of workspaces;
• wherein aggregating said measure column vector according to said plurality of grouping column vectors includes aggregating, in each workspace of said plurality of workspaces, one or more codes of said measure column vector that correspond to an actual grouping key mapped to a dense identifier corresponding to said each workspace;
• wherein, for a first row of said measure column vector and said plurality of grouping column vectors, aggregating said one or more codes includes:
• generating a particular grouping key based on codes in said plurality of grouping column vectors;
• determining that said particular grouping key is mapped to a particular dense identifier;
• and
• updating a particular workspace that corresponds to said particular dense identifier based on a particular code of said measure column vector.