Hi Amparo,
What is the LONGEST string length of QLABEL in your dataset?
Just look for the longest string for QLABEL, and then count the number of characters without the blanks at the end. That should also be the field length.
The reason for this rule is that the FDA still uses the outdated SAS-XPT (Transport 5) format, which is "fixed field length", i.e. for each variable, every field in the dataset has the same length for each row, and shorter strings (than the longest one) are completed with blanks.
The FDA wants to minimize file sizes, and that is why they developed this rule.
Essentially, this rule should not exist, as at the FDA, datasets should be stored in databases (of whatever type). Modern databases take care of optimizing size, do indexing for fast access etc.. Most of the FDA reviewers however still use files "as is", sometimes even working with memory sticks ...
Hi Amparo,
Based on your screenshots, I assume that a maximum length of values in QLABEL variable of SUPPPS dataset is 31.
Try set a length for QLABEL variable as 31 Chars. For example, in SAS: "length QLABEL $31;"
Kind Regards,
Sergiy
Dear Amparo,
What Sergiy means it that you should set the variable length for QLABEL to 31 when GENERATING the dataset.
After that, do NOT forget to check whether all variable values are correct and none is truncated. Do this using the "SAS System Viewer" or the "SAS Universal Viewer" (it looks as you used the latter).
With the use of a modern format, like XML, JSON or RDF, we should not have such issues ....
Hi Jozef,
Thank you for your reply. I found there is a "blank" at the end of string therefore program detect it as 32 characters but validator check as 31. The issue had been solved after removing the blanks.
Best regards,
Amparo
Hi Sergiy,
Thank you for explanation, we found there was a "blank" at the end of string. the issue had been solved after removing the "blank".
Best regards,
Amparo
Hi,
We have an issue about SD1082/FDAB018. We use Pinnacle21 community version 3.0.2 to validate our SEND datasets (configuration as IG 3.0) and get errors about "Variable length is too long for actual data". No matter what we do to the value of QLABEL (re-sized the length) the report still give us same result (please find snapshot in the attachment). or is there any we misunderstanding?
thank you
Best regards,
Amparo