For a particular query with pubmed (through bookends) I get 588 matches and bookends asks me how many should I download— I say 100, it downloads only 40!
Next for the same query I say get em ALL, it downloads some 376. What am I missing (besides the references I may be missing)?
(Toggling 'Do not clear reference list between searches' does not change or fix this behaviour)
Jon wrote:I'm seeing that PubMed is generating a lot of
"Error occurred: Document retrieval error: document does not exist"
errors, which accounts for the discrepancy. So this looks like a problem on their end.
Jon
Sonny Software
Thanks for the reply Jon.
But, when I ask BE to fetch 100, why does it stop at 40 when 376 non-error hits are available for download when I select 'ALL'?
Jon wrote:I have no idea -- the problem is at PubMed, not here. You'll have to ask NLM.
BTW, it's not "stopping" at 40 -- it tries to get all 100, but can't find 60 of the 100, not necessarily the last 60.
Jon
Sonny Software
I am suspecting these 'error's are created due to PMIDs being allocated to e-publications ahead of print or future content that is beginning to be indexed, but not fully indexed yet? (when I ask for hits only from 2003 or 2004 I seem to get all the hits as expected)
I guess til NLM fixes it, I just have to work with the awareness that the no. of references I get could be a lot fewer than expected.