I might be wrong here (someone please correct me if I am) but you can only have multiple main tables if there is a 1:M relation between them - i.e. every B includes all the fields of A plus some extra fields, some of which make B unique when looked at in combination with the fields inherited from A. Dimension tables that refer to A must include all the key columns for A, and those that refer to B must refer to all the key columns for B (which by definition will include all the key columns that B inherited from A).
Could you check that this is the case? cheers, Richard On Wed, 2006-06-28 at 11:46 +0200, allank wrote: > Hi all, > I am unable to query from a biomart schema having two tables and (at the > moment) one dimension table, I get a " SQL FAILED - LIKELY XML > MISCONFIGURATION ERROR ..." error at the result display stage on the > biomart web application. > I have setup the biomart table structure in accordance with the multiple > main table requirements, in Martj I "validate all" and I get no errors, > when using the web application, I am able to go smoothly through the > filter process and the field-section-for-output stages the pane at the > right indicates I have 32 entries at the output level, when I attempt to > proceed to the output display page, I get the message " SQL FAILED - > LIKELY XML MISCONFIGURATION ERRORUnknown column 'main.timepoint_id_key' > in 'where clause'". > Could the error have been caused from using a key field (in my case > timepoint_id_key) for fields from the dimension table that is not found > in all the main tables. If this is so, how can I manage the situation > where I have a main table (dataset__A__main) having a one to many > relationship with one dimension table (dataset__C__dm) and other main > table (dataset__B__main) which will have one to many relationship with > several dimension tables. > > Allan. > -- Richard Holland (BioMart Team) EMBL-EBI Wellcome Trust Genome Campus Hinxton Cambridge CB10 1SD UNITED KINGDOM Tel: +44-(0)1223-494416
