Histograms axes limits are wrong when using geom_smooth() about ggextra

[daattali]

Thanks for the report. Fixing

[WilDoane]

A version of this problem appears to still exist when geom_smooth() extends the axis range.

gg <-
    ggplot(mtcars, aes(x = mpg, y = cyl)) + 
    geom_point() 
ggMarginal(gg, type = "histogram")

gg <-
    ggplot(mtcars, aes(x = mpg, y = cyl)) + 
    geom_point() +
    geom_smooth()
ggMarginal(gg, type = "histogram")

package       * version date       source        
 colorspace      1.2-6   2015-03-11 CRAN (R 3.2.2)
 devtools        1.10.0  2016-01-23 CRAN (R 3.2.3)
 digest          0.6.9   2016-01-08 CRAN (R 3.2.3)
 ggExtra       * 0.3.1   2015-11-07 CRAN (R 3.2.3)
 ggplot2       * 2.1.0   2016-03-01 CRAN (R 3.2.3)
 gridExtra       2.2.1   2016-02-29 CRAN (R 3.2.3)
 gtable          0.2.0   2016-02-26 CRAN (R 3.2.3)
 labeling        0.3     2014-08-23 CRAN (R 3.2.2)
 memoise         1.0.0   2016-01-29 CRAN (R 3.2.3)
 munsell         0.4.3   2016-02-13 CRAN (R 3.2.3)
 plyr            1.8.3   2015-06-12 CRAN (R 3.2.2)
 Rcpp            0.12.3  2016-01-10 CRAN (R 3.2.3)
 RevoUtilsMath * 3.2.3   2016-02-09 local         
 scales          0.4.0   2016-02-26 CRAN (R 3.2.3)

[daattali]

Thanks for the report. I'm really busy for the next 2 months, I'll try to take a look when I can, but unfortunately this is low priority at the moment

[daattali]

@WilDoane I looked into this a bit but wasn't able to find a good solution. What I tried to do is to just copy the scale aesthestics from the main plot to the marginal plot, but for some reason it doesn't work. If anyone wants to tackle this, I'd gladly welcome a PR.

[maelle]

@daattali "What I tried to do is to just copy the scale aesthestics from the main plot" -> where exactly in ggMarginal? Do you mean

        ggplot2::scale_x_continuous(breaks = mean(getLimits(pbTop, "x")),
                                    labels = ylabel)

?

I'm looking at the problem because a colleague of mine uses custom x and y limits.

[daattali]

@masalmon getLimits() and getScale() are two functions I use to try to match the original plot. If your colleague uses custom x and y limits, she should be fine, that works. Example:

ggExtra::ggMarginal(
  ggplot(mtcars, aes(mpg,wt)) +
    geom_point() + 
    scale_x_continuous(limits = c(-5, 60))
)

What doesn't work is using geom_smooth() - then the y-axis marginal plot is a bit out of place

[jimoeppen]

The original issue raised by zakahmad in 2015 still seems to be present, without using geom_smooth. If the default axis limits are over-ridden, the x-axis marginal plot aligns correctly with the data, but the y-axis marginal plot does not. Here is some code to illustrate the situation [R 3.4.3; ggExtra_0.7; ggplot2_2.2.1]:

df <- data.frame(x = rnorm(500, 50, 10), y = runif(500, 0, 50))
p1 <- ggplot2::ggplot(df, ggplot2::aes(x, y)) + ggplot2::geom_point() +
scale_x_continuous(limits = c(0, 150)) +
scale_y_continuous(limits = c(0, 100))
ggMarginal(p1, type = "histogram")

[crew102]

@jimoeppen , the problem with using limits with ggMarginal that you mention is related to #81.

[jimoeppen]

Thank you for the rapid response, and for all the work that goes into ggExtra. I think #81 implies that setting limits on x affects the y marginal, and vice-versa. As far as I have seen, the x marginal responds correctly to the x limits. It is only the y marginal that does not reflect user-specified y axis limits. Best wishes Jim Oeppen Associate Professor Biodemography Unit Department of Public Health University of Southern Denmark (SDU) J.B. Winsløws Vej 9B DK-5000 Odense C Office Phone: [phone] From: Chris Baker [mailto:[email protected]] Sent: 11 January 2018 16:39 To: daattali/ggExtra <[email protected]> Cc: Jim Oeppen <[email protected]>; Mention <[email protected]> Subject: Re: [daattali/ggExtra] Histograms axes limits are wrong when using geom_smooth() (#9) @jimoeppen<https://github.com/jimoeppen> , the problem with using limits with ggMarginal that you mention is related to #81<#81>. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#9 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AHsb6IHW57spF4DPz_nRSeeF-sYDRX5Jks5tJisbgaJpZM4EdDzn>.

[crew102]

Hi @jimoeppen, both axis are actually problematic. You can see both graphs in #81 for this, or just check out the below code, in which the x marginal plot is wrong and we don't even get a y marginal plot.

library(ggplot2)
library(ggExtra)

p <- ggplot(data = mtcars) +
  geom_point(aes(wt, mpg)) +
  ylim(c(25, 35))

ggMarginal(p, type = "histogram")

The bigger issue mentioned in #81 is that ggMarginal is not using the correct range for the marginal plots.

[jessiepluto]

I just encountered this same issue -- ggMariginal histograms are not lining up with the y-axis correctly after adjusting the limits. Is there any update on a solution?

[daattali]

A huge feature was just added, the fix for this is next

[daattali]

fixed