diffobj
diffobj copied to clipboard
Rethink Default Mode Selection
Not entirely satisfied how we chose side by side vs. unified. Here is an example where unified would have made more sense
< .ref > .new
@@ 1,6 @@ @@ 1,6 @@
intercept slope intercept slope
< -1717.000000 101.000000 > -3.541306e+13 7.012486e+11
rsq rsq
< 0.938679 > 9.386790e-01
attr(,"class") attr(,"class")
[1] "fastlm" [1] "fastlm"
vs.
< .ref
> .new
@@ 1,4 / 1,4 @@
intercept slope rsq
< -1717.000000 101.000000 0.938679
> -3.541306e+13 7.012486e+11 9.386790e-01
attr(,"class")
[1] "fastlm"
Need some metric that computes percentage of lines that need to be wrapped or some such, but this may require two passes.