Forums: Troubleshooting and Problems
I think it is always wise to have a second option to edit/correct define.xml files that you automatically generate from things like Excel worksheets. There are several good tools on the market for editing/correcting define.xml files that are very user-friendly (with wizards and including validation) allowing you to take control. And of course you can always learn about define.xml in detail (including the XML) e.g. by following a by CDISC organized training. As a last resource, you can also correct your define.xml files using a normal XML editor. Some are free, others cost less than 100-200$. We do use such one in the CDISC define.xml trainings.
In the case that you have, correcting the define.xml would take something like 3 minutes or less using an XML editor.
Jozef Aerts
CDISC Define.xml trainer
So would I be correct in saying that there are limitations in the Pinnacle21 converter, which cannot be overcome (or are prohibitively difficult to overcome) using Excel alone? We're trying to figure out if we're doing something wrong, or if this is just the state-of-play of the conversion tool.
We're working on producing a define.xml from from a spec spreadsheet in Excel, using the Pinnacle21 system. We're noticing that for value-level variables, the where clause will be duplicated if the same text is used to qualify two or more variables. In the attached example, DRVMTH and one other variable are both qualified by PARAMCD, which is normal, but the where-clause gets repeated. It is worth noting that we have checked that the WhereClause ID is definitely distinct. We can resolve this issue if we modify the where clause value to be distinct for each variable we're "value-leveling" by including the variable label, rather than the value decode (in this case Sum of Diameters (mm)). However, I think it makes more sense to have the decode in the define.xml, and shouldn't these be treated as separate clauses based on the WhereClause ID, solely? Otherwise, what's the point of having an ID if it doesn't uniquely identify the clause?
Cheers,
David Oakley
Attached Files