Timeline for R: Box-plot on log scale vs. log-transforming *then* creating box-plot: Don't get same result
Current License: CC BY-SA 3.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 9, 2016 at 20:13 | vote | accept | Meg | ||
| May 8, 2016 at 8:06 | comment | added | Hao Ye | @Meg I think it would depend on whether you think the whiskers (representing the range) should be the same length in log-space or unscaled. Given that you are log-transforming the data, the former probably makes more sense. | |
| May 7, 2016 at 22:02 | comment | added | Meg | Thank you. It's not clear to me, then, what's the "most correct" way to represent the data. This post (which is for Stata, not R, but discusses the same issue) seems to imply the best practice would be to first log-transform the data, then manually re-label the box-plot in terms of the original units: stata.com/support/faqs/graphics/… Thoughts on this approach? | |
| May 7, 2016 at 16:15 | history | answered | Horst Grünbusch | CC BY-SA 3.0 |