For the purpose of QC of concentration data released by data management, I need to compare these data with the raw concentration data available from the biolaboratory. I’m thinking to use Phoenix WNL for such purpose.
However, when preparing data from the biolab, subject is only identified on the first row of each subject’s data, as in the example below:
Subject Period Time Concentration
01 1 0.00 BQL
1 0.25 0.107
1 0.50 4.12
2 0.00 BQL
2 0.25 0.254
2 0.50 3.43
02 1 0.00 BQL
1 0.25 2.163
1 0.50 8.60
2 0.00 BQL
2 0.25 0.872
2 0.50 4.38
I need to copy the ID of the subject for all the rows, as below
Subject Period Time Concentration
01 1 0.00 BQL
01 1 0.25 0.107
01 1 0.50 4.12
01 2 0.00 BQL
01 2 0.25 0.254
01 2 0.50 3.43
02 1 0.00 BQL
02 1 0.25 2.163
02 1 0.50 8.60
02 2 0.00 BQL
02 2 0.25 0.872
02 2 0.50 4.38
Does forum has any suggestion on the use of Phoenix functions?
I beleive the request for a lag function or similar that would enable this is logged as QC 11694, I’ll drop a note to remind the Product Manager to look at this and/or suggest a workaround within Phoenix.
I am thinking even an option on import to carry forward a value until it is next set could be a useful improvement that might be relatively quick to implement.
nice to meet you in cyberspace. I hope you survived “Trumpistan” last month.
I think that drag/drop in Excel is prone to errors. IMHO, a lag-function (Simon’s QC 11694) is hardly better. What if you have complete missings (vial broken in centrifugation, no result at all, blahblah)?
I would say the only solution which does not require a lot of QC-work (do it twice, 100% data check, etc) is to politely but assertive ask the lab to provide complete data. The samples are labeled with all the required information which is carried through all sample processing steps until it ends up in the CDS. Why the heck do these guys/gals leave you out in the rain with incomplete data?
They should revise their data transfer procedure. Makes not only you happy but the life of inspectors easier. I don’t expect that they will like any ad hoc “solution”. Good luck writing an SOP for it. Opens a can of worms. I would refuse to work with such a pile of pooh. I provide the lab before with my required data format. Yep, CSV – not bloody Excel. I also require a checksum. If the format is not correct or no checksum is provided, I ring up the sponsor.
I agree - would be better to get source correct/completed in the long run - but I also want people to be able to work around some data format/layout problems themselves! Happy New Year Helmut. Simon.
