Support non-matching value attributes for merges
backlog
T
Technological Ostrich
In the Q merge tool, you can merge data that has different value attributes and pick which dataset to use to get the labels for the values. Especially with trackers where survey systems (and questions) can change over time, it’s imperative this is supported by our new merge tools in Displayr so customers don’t have to jump through hoops and spend TONS of time to try to reformat data to be merged. In this specific case, the user was combining a survey that was done across countries/languages.
J
Jamie Morris
backlog
J
Jamie Morris
Merged in a post:
Less restrictions on merging variable data
I
Improved Chimpanzee
Less restrictions on merging data from variables. i.e. the variables can be within a variable set, they can be Nominal-Multi (we can check to be sure the code frame or values are the same) or Text, variable labels don't have to be in the same order for each.
The workaround in this case was to convert the Nominal-Multis to Binary-Multis, then
Combine > As a Variable Set > Merge Variables
, then convert back to the Nominal-Multi and relabel the values. This solution is not obvious to the user.J
Jamie Morris
under review
J
Jamie Morris
Merged in a post:
Improve combine data sets functionality in Displayr
X
Xenacious Cat
When users rely on third providers to provide data sets for trackers, they don’t always have a say in the format or naming and labelling. This can lead to many inconsistencies that make the current merging setup not as user friendly.
Users often want to be able to compare data sets prior to merging in order to see how they align. It’s not always intuitive to see what needs to be manually merged and what rules to use, especially when the function essentially merges first before indicating any matching.