Correction for the detection of duplicated emofs with no measuremenTypeID

[cyrilrader]

Removed this filter '!is.na(measurementTypeID' so the emof with no measurementTypeID are not excluded from the check. Added the column measurementType to the 'select' and 'group_by'.

Changes have been tested on my fork of the tool first and with this modified dataset:
test_modified_biofun.zip

[rubenpp7]

Hi @cyrilrader

I feel that this commit is replacing one functionality by another and it may introduce a bug.

For example:

• cases where measurementTypeID exist AND it's the same for 2 rows within the same eventID&occurrenceID BUT where the measurementType field is populated and different for those 2 rows will NOT be detected anymore.

I thought that the issue we wanted to solve was to detect exact duplicated rows, why not create a different check that looks at all the fields, instead of only at one more field?

[cyrilrader]

Hi @rubenpp7

Yes, you are right. In the case you described, the duplicates won't be detected. But, it would be the same for a check that looks at all the fields no ?
I'll try to find another solution that solve this

[rubenpp7]

@cyrilrader exactly, that's why the point would be to add a new check instead of replacing an already existing one