Extended database engine providing versioning and embedded analytics

1. A method comprising:

storing in a database one or more tables including records with a plurality of fields;

retrieving the records from the database and storing data from the records in a cache in order to perform a calculation, the records being rearranged so that each record has an identifier and at least for one record multiple versions of that record are stored in the cache, the different versions of the record including one or more variations in field values, wherein the records are stored in the database with groups of fields organized into at least two field groups with at least two fields each; and

selectively retrieving from the database and storing in the cache one or more, but not all, field groups on an as-needed basis to perform a calculation.

2. The method of claim 1 , wherein the cache is in a server separate from the database.

3. The method of claim 1 , wherein, for the versions maintained in the cache with one or more of the field groups, to retrieve data in a first field in response to a query, retrieving the field group that includes the first field without retrieving other field groups.

4. The method of claim 1 , further comprising maintaining an inverse field set such that for each value of a field, a set is created of identifiers that may have the same field value.

5. The method of claim 1 , wherein retrieving the records from the database includes storing the data in an input data cache with a record page with an identifier for each record and wherein there is a set of versions for at least some of the records, such that a record with multiple versions has one identifier.

6. The method of claim 5 , further comprising maintaining an inverse set for a plurality of fields, the inverse set including, for each field, a list of identifiers of records that have the same value for that field, including fields in records for which there are multiple versions.

7. A system comprising:

a database having one or more tables including records with a plurality of fields; and

a cache memory for storing data from the records, the data being rearranged so that each record has an identifier and at least one record has multiple versions of that record stored in the cache, the different versions of the record including one or more variations in field values;

wherein the records are stored in the database with groups of fields organized into at least two field groups with at least two fields each, wherein records are selectively retrievable from the database and stored in the cache such that one or more, but not all, field groups are retrievable on an as-needed basis to perform the calculation.

8. The system of claim 7 , wherein the versions are maintained in the cache in field groups, such that in order to retrieve data in a first field in response to a query, the system retrieves the field group that includes the first field without retrieving other field groups.

9. The system of claim 7 , wherein the cache memory has an inverse field set such that for each value of a field, a set is created of all identifiers that may have that field value.

10. The system of claim 7 , wherein the data is stored in an input data cache with a record page with a list of identifiers such that each identifier identifies a record, and sets of versions are each associated with a single identifier.

11. The system of claim 10 , wherein the cache memory has an inverse set for a plurality of fields, the inverse set including the field and a list of identifiers of records that have the same value for that field.

12. The system of claim 7 , wherein the cache memory is in a server separate from a database system that has the database tables.

13. The system of claim 7 , wherein the cache memory is in a database system that has the database tables.