[R] how to change the y-axis to logarithmic in a barplot ggplot

[Kimmo Elo]

Hi,

Have you tried:

scale_y_continuous(trans='log10')

HTH,

Kimmo
________________________________________
Lähettäjä: Maria Lathouri via R-help <r-help using r-project.org>
Lähetetty: maanantai 17. heinäkuuta 2023 23.01.57
Vastaanottaja: Maria Lathouri via R-help; Ivan Krylov
Aihe: Re: [R] how to change the y-axis to logarithmic in a barplot ggplot

Hi Ivan,

yes you are right. I have many values close to zero and if I use log, they are instantly transformed to negative values and then the bars show above. The sqrt trans worked ok but when I used the pseudo_log_trans, it did the work!

Many thanks.

Kind regards,
Maria






Στις Δευτέρα 17 Ιουλίου 2023 στις 04:43:59 π.μ. GMT+1, ο χρήστης Ivan Krylov <krylov.r00t using gmail.com> έγραψε:





On Sun, 16 Jul 2023 21:51:10 +0000 (UTC)

Maria Lathouri via R-help <r-help using r-project.org> wrote:

> As you can see, the values range from 0 to 400. I want it to plot it
> in bars; when I am plotting it as you can imagine the values near
> zero don't show at all.


A logarithmic scale won't work when the range of numbers to display
includes 0. It would have worked if it was some small non-zero
quantity, but an actual zero is infinitely far down; it won't fit.

There is a number of transformations described in
help(scale_y_continuous). You may find trans="pseudo_log" or
trans="sqrt" useful.

help(geom_bar) recommends geom_col() instead of geom_bar() to represent
values (not counts of occurrences).

We could help you more efficiently if you clarified the relationship
between your data (with columns `ID`, `values`, `databases`) and your
code (which mentions `Temp`, `mean`, `Glass`).

--
Best regards,
Ivan

______________________________________________
R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.