Closed jwang-lilly closed 1 year ago
Hi Jiang,
unfortunately, I have only implemented >=
and <=
.
The difference between >
and >=
matters only with few unique values (as in the example suicide
dataset) and if use_midpoints = FALSE
.
For that reason and because this feature was not requested thus far, I have not implemented it and also do not plan to do so – unless there is a good argument why it is needed. You are mainly interested in comparing the results to a different tool, right?
So you can unfortunately not set direction = ">"
, but you could try setting use_midpoints = TRUE
which should bring the optimal cutpoint closer to the one that would be obtained with direction = ">"
and use_midpoints = FALSE
.
(I think with use_midpoints = TRUE
the optimal cutpoint would actually be identical with >
and >=
)
Best, Christian
Thanks very much Christian. This is very helpful!
Dear Dr. Thiele @Thie1e
We have a quick question about allowable inputs for parameters of cutpointr(). direction currently only accepts" >=" or "<=" but not ">" or "<". We are trying to match the cutpointr output with output from another tool and we figured out the difference in the ">=" in cutpointr and ">" in the other tool. Is there any way we can set ">" in cutpointr()? Any advice is greatly appreciated.
Thanks so much for your help.