Repository navigation
all.equal only reports differences in first column with differences #7903
Description
Activity
This is documented in the Details section of
?all.equal.data.table:For efficiency data.table method will exit on detected non-equality issues, unlike most all.equal methods which process equality checks further.
all.equal.data.tablecould state so in its output (either in the return value or as a message).We should offer an option for computing the full differences if requested
all.equal.data.tablecould state so in its output (either in the return value or as a message).I would say it is safe to assume that user will read manual of a function that he/she is using, especially if it "misbehaves" for them.
Also there are couple open issues regarding all.equal, one pointing to rewrite it. #7816 (comment) that could possibly address this issue.
I agree with @jangorecki that I should have read the documentation more carefully. However, I did read the documentation and somehow I didn't connect For efficiency data.table method will exit on detected non-equality issues, with the single output printed.
Then we could improve documentation at the very least. Thanks for pointing that out.
I would say it is safe to assume that user will read manual of a function that he/she is using, especially if it "misbehaves" for them.
I'm not sure that's safe. Regardless, it's still not good UX.
When comparing two data.tables with differences in multiple columns only the difference in the first column is printed.
See the code below:
I expected similar output printed as when using all.equal on two data.frames, i.e. both columns that have differences reported. The latter is also what is described as output value for the function.