1
$\begingroup$

Consider the code

dataBC = 10^(RandomReal[6, 10]);
dataTotal = dataBC = 10^(RandomReal[6, 10]);
BoxWhiskerChart[{dataBC, dataTotal}, {"Mean"}, ScalingFunctions -> "Log"]

enter image description here

Is it possible to force the y-axis ticks to be labeled in the form $10^x$ so that, for example, $100$ and $1000$ are shown as $10^2$ and $10^3$, respectively?

$\endgroup$
5
  • $\begingroup$ Search the site for "scientific notation" and "axis". You would find Writing axis into scientific notation, Scientific notation for the axes of a Plot. Those would also lead to About the number format in ticks, which is probably most directly relevant to you. $\endgroup$
    – MarcoB
    Commented Apr 27, 2023 at 15:07
  • $\begingroup$ 241982. $\endgroup$
    – Syed
    Commented Apr 27, 2023 at 15:10
  • 2
    $\begingroup$ log box-whisker plots tend to deceive the reader into thinking there's a lot more probability mass in the lower quartile and a lot less in the top than there really is. $\endgroup$
    – flinty
    Commented Apr 27, 2023 at 15:22
  • $\begingroup$ @flinty I see. Do you have any better suggestion to show data that contains different orders of magnitudes? $\endgroup$
    – sam wolfe
    Commented Apr 27, 2023 at 17:10
  • $\begingroup$ @samwolfe I'd group the data according to order of magnitude and only look at aggregates per group, or I'd try to nest the data in an interactive visualization like a tree map, but there's generally little hope for good visual depictions of such data. In this case you have to ask yourself what benefit a visualization even brings. The trouble with taking a mean (like the middle bar) across 5 orders of magnitude is that it's very much 'mean'ingless. The variance of the data is far more significant here than the mean - "Never cross a river that is on average 4 feet deep!" $\endgroup$
    – flinty
    Commented Apr 27, 2023 at 18:27

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.