csdaw / ggprism Goto Github PK

I generate a lot of graphs of the same type. And they all have a different scale along the y-axis. And the step.increase default value is not suitable, because the lines on the graph with the p-value stick together. If I change the step.increase standard value, then the lines, on the contrary, diverge too much.

stat.test <- total %>% rstatix::wilcox_test(as.formula(paste(col_names[col_i], "group", sep="~" ))) %>% rstatix::add_y_position()
stat.test$y.position[3] = stat.test$y.position[1]

Hi!

i chanced upon your package and was wondering how i can add new themes to it?

Thank you very much and look forward to hearing back from you.

Need to release a patch so p-value vignette passes CRAN checks. Otherwise ggprism will get removed from CRAN 😵

To do:

• Fix p-values vignette
• Increase test coverage

Prepare for resubmission:

• Knit README.Rmd
• R CMD check
• devtools::check(remote = TRUE, manual = TRUE)
• Update NEWS.md
• Push final changes to trigger GitHub Actions
• Review GitHub Actions results
• devtools::check_win_devel()
• rhub::check_for_cran()*
• Update cran-comments.md
• Push

*I used:

rhub::check_for_cran(
  platforms = c("fedora-clang-devel",
                "windows-x86_64-oldrel"),
  env_vars=c(
    `_R_CHECK_FORCE_SUGGESTS_` = "true",
    `_R_CHECK_CRAN_INCOMING_USE_ASPELL_` = "true", 
    R_COMPILE_AND_INSTALL_PACKAGES = "always"
  )
)

Submit to CRAN:

• usethis::use_version("patch")
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()
• Tweet

Small release mainly to preserve compatibility with ggplot2 v3.4.0 (but also fixes a bug and adds a small feature, see NEWS.md).

Prepare for submission:

• Update NEWS.md
• url_checker::url_check()
• R CMD check
• devtools::check(remote = TRUE, manual = TRUE)
• revdepcheck::revdep_check(num_workers = 4)
• Push final changes to trigger GitHub Actions
• Review GitHub Actions results

Optionally fix things that need fixing, then repeat the above.

Submit to CRAN:

• devtools::check_win_devel()
• rhub::check_for_cran()*
• Update cran-comments.md
• usethis::use_version("patch")
• devtools::build_readme()
• Commit
• devtools::check(remote = TRUE, manual = TRUE)`
• Push again
• Review GitHub Actions results
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()

*I used:

rhub::check_for_cran(
  platforms = c("fedora-clang-devel", "windows-x86_64-oldrel"),
  env_vars=c(
    `_R_CHECK_FORCE_SUGGESTS_` = "true",
    `_R_CHECK_CRAN_INCOMING_USE_ASPELL_` = "true", 
    `R_COMPILE_AND_INSTALL_PACKAGES` = "always"
  )
)

Prepare for release:

• devtools::build_readme()
• R CMD check
• devtools::check(remote = TRUE, manual = TRUE)
• Push final changes to trigger GitHub Actions
• Review GitHub Actions results
• devtools::check_win_devel()
• rhub::check_for_cran()*
• Update cran-comments.md and push

*I used:

rhub::check_for_cran(env_vars=c(`_R_CHECK_FORCE_SUGGESTS_` = "true",
  `_R_CHECK_CRAN_INCOMING_USE_ASPELL_` = "true", 
  R_COMPILE_AND_INSTALL_PACKAGES = "always"))

Submit to CRAN:

• usethis::use_version("patch")
• devtools::submit_cran()
• Approve email

Wait for CRAN...

Need to fix issues and resubmit.

Need to release a patch so tests won't fail when ggplot2 v3.4.0 is released.

Prepare for resubmission:

• devtools::build_readme()
• url_checker::url_check()
• R CMD check
• devtools::check(remote = TRUE, manual = TRUE)
• Update NEWS.md
• Push final changes to trigger GitHub Actions
• Review GitHub Actions results
• devtools::check_win_devel()
• rhub::check_for_cran()*
• Update cran-comments.md
• Push

Submit to CRAN:

• usethis::use_version("patch")
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()
• Tweet

*I used:

rhub::check_for_cran(
  platforms = c("fedora-clang-devel", "windows-x86_64-oldrel"),
  env_vars=c(
    `_R_CHECK_FORCE_SUGGESTS_` = "true",
    `_R_CHECK_CRAN_INCOMING_USE_ASPELL_` = "true", 
    `R_COMPILE_AND_INSTALL_PACKAGES` = "always"
  )
)

Hi @csdaw,

I am using the add_pvalue function, and I am wondering whether it would be possible to have the P values nicely formatted as P = 1x10^-10 rather than P=1e-10. For doing this, I usually use latex2exp::TeX but I don't understand how to integrate it into the add_pvalue function.

I am now using rstatix to both perform the statistical test and create the p-value table with the correct x and y position:

p <- ggplot(ToothGrowth, aes(x = factor(dose), y = len)) +
  geom_boxplot(aes(fill = dose), colour = "black") +
  theme_prism() +
  theme(legend.position = "none")
stat.test <- rstatix::t_test(ToothGrowth, len ~ dose, ref.group = "0.5") %>%
  rstatix::add_xy_position()

My first thought was to create a new field in stat.test to store the LaTeX expressions generated by latex2exp::TeX, as follows:

stat.test$p.latex <- latex2exp::TeX(stat.test$p.latex)

where p.latex Is a string vector containing LaTeX expressions (let's say "1 \\times 10^{-10}") which I generated with another function.
However, the stats.test object is not able to store expressions (as those outputted by latex2exp::TeX).

I then tried to modify the

p + add_pvalue(stat.test, label = "p = {p.latex}")

bit and the add_pvalue function to have the expression printed, but I have no idea on how to do this.

Would you be able to help me?
Many thanks in advance!

Hi

We are preparing the next version of ggplot2 and our tests show that ggprism is having issues with it. Specifically we now store the computed params in computed_geom_params and computed_stat_params. This means that these are the field you should be expecting in your tests if you feel it necessary to peek into the ggplot object.

The failing tests are:

----- FAILED[data]: test-annotation_ticks.R<66--66>
call| expect_true(g1$layers[[2]]$geom_params$colour == "red")
diff| Expected TRUE, got 'logical' of length 0
----- FAILED[data]: test-annotation_ticks.R<69--69>
call| expect_true(g2$layers[[2]]$geom_params$colour == "red")
diff| Expected TRUE, got 'logical' of length 0

where you should look at g1$layers[[2]]$computed_geom_params$colour instead

we plan to release the next version of ggplot2 on May 31st

Hey there,

I am currently running the following R code:

p_vals <- tibble::tribble(
  ~group1, ~group2, ~p.adj,   ~y.position,
  "audi",   "dodge",     0.01 , 3.99,
  "audi",   "chevrolet",  0.01,  3.5
)

mpg %>%
  mutate(displ = as.numeric(displ)) %>% 
  ggplot(aes(x = manufacturer, y = displ, fill = manufacturer))+
  geom_bar(position=position_dodge(), stat="identity") +
  geom_jitter(height = 0, width = 0.2, size = 4) +
  ylim(0,4) + 
  theme_prism(base_size = 16)+
  scale_fill_manual(values = cbbPalette)+ 
  theme(legend.position = "none") +
  add_pvalue(p_vals, label = "p = {p.adj}", tip.length = 0, label.size = 4)

And receiving the following error:

Error:
! Problem while computing aesthetics.
ℹ Error occurred in the 3rd layer.
Caused by error in `FUN()`:
! object 'manufacturer' not found
Backtrace:
  1. base (local) `<fn>`(x)
  2. ggplot2:::print.ggplot(x)
  4. ggplot2:::ggplot_build.ggplot(x)
  5. ggplot2:::by_layer(...)
 12. ggplot2 (local) f(l = layers[[i]], d = data[[i]])
 13. l$compute_aesthetics(d, plot)
 14. ggplot2 (local) compute_aesthetics(..., self = self)
 15. base::lapply(aesthetics, eval_tidy, data = data, env = env)
 16. rlang (local) FUN(X[[i]], ...)
Error in (function (mapping = NULL, data = NULL, stat = "bracket", position = "identity", :

ℹ Error occurred in the 3rd layer.
Caused by error in `FUN()`:
! object 'manufacturer' not found

When I run the same code without the add_pvalues() command, it creates the plot. I can also add pvalue lines with geom_signif(), but those stay small / don't get larger with the theme_prism(base_size=16) argument so I'd like to make the add_pvalues() arg work .

Thanks for any advice! Not sure how to fix this one~

I tried to add the legend title to my data, but I found that if I added theme_prism(base_size = 14) my legend title would disappear, if I commented this out the legend title would appear again, I don't know what is wrong, can someone help me?
I used R4.1.3 and downloaded prism version 1.0.2

ggprism

Things to include in the future:

• Prism .pzfx file with examples and a vignette recreating them with this package. Or even a pkgdown site...
• A vignette which downloads data from figshare or whatever and recreates the figures from a paper. Could also be a page on a future pkgdown site...
• Examples which use packages that will go well with ggprism e.g. kassambara/rstatix, eclarke/ggbeeswarm

Other packages

Some features/functions of other packages would fit in ggprism, but I don't want to extend the dependencies of ggprism beyond those of ggplot2. So need to rewrite them into ggprism.

• Significance annotations. Combine geom_signif from const-ae/ggsignif and stat_pvalue_manual from kassambara/ggpubr
• Some more axis guides. See stefanedwards/lemon
• Repeat facet axes. See stefanedwards/lemon

Hi,
I've got this error when I updated the ggplot2 to the latest version 3.4.4.9000. The legend.text.align theme element is not defined in the element hierarchy.

I think it has something to do with ggplot2 deprecating the legend.text.align argument of theme().

Thanks!

Fixes compatibility with ggplot2 v3.5.0.

Prepare for submission:

• Update NEWS.md
• urlchecker::url_check()
• R CMD check
• devtools::check(remote = TRUE, manual = TRUE)
• revdepcheck::revdep_check(num_workers = 4)
• Push final changes to trigger GitHub Actions
• Review GitHub Actions results

Optionally fix things that need fixing, then repeat the above.

Submit to CRAN:

• devtools::check_win_devel()
• rhub::check_for_cran()*
• Update cran-comments.md
• usethis::use_version("patch")
• devtools::build_readme()
• Commit
• devtools::check(remote = TRUE, manual = TRUE)`
• Push again
• Review GitHub Actions results
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()

*I used:

rhub::check_for_cran(
  platforms = c("fedora-clang-devel", "windows-x86_64-oldrel"),
  env_vars=c(
    `_R_CHECK_FORCE_SUGGESTS_` = "true",
    `_R_CHECK_CRAN_INCOMING_USE_ASPELL_` = "true", 
    `R_COMPILE_AND_INSTALL_PACKAGES` = "always"
  )
)

Hi

We preparing the next release of ggplot2 and our reverse dependency checks show that your package is failing with the new version. Looking into it we see that your package somehow includes references to ggplot2 code from when it was build, either by reloading saved plot objects or by grabbing functions at build time. Please see https://www.tidyverse.org/blog/2022/09/playing-on-the-same-team-as-your-dependecy/ for some more info about why this can cause issues.

You can install the release candidate of ggplot2 using devtools::install_github('tidyverse/[email protected]') to test this out.

We plan to submit ggplot2 by the end of October and hope you can have a fix ready before then

Kind regards
Thomas

Hi first of all great package! However I have an issues with hiding the legend....adding:

plot + theme(legend.title = element_blank()) doesnt seem to work....adding show.legend = F to the aes of ggplot does work but every time I add plot + theme_prism() the legend pops back up

This is an option in stat_pvalue_manual to hide insignificant p values from being plotted and seems to be missing from add_pvalue function. Would you be able to add it?

Thank you for your great work.
I found a problem with the label.
The problem is that;
An "i" in the label string is forced to be capitalized.

iris %>% ggplot(aes(x = Sepal.Length, y = Sepal.Width, color = Species))+ geom_point()+ xlab("all small iiiiis")+ theme_prism()

iris %>% ggplot(aes(x = Sepal.Length, y = Sepal.Width, color = Species))+ geom_point()+ xlab("all small iiiiis")+ theme_bw()

my environment is;
Mac mini (M1)
R version 4.4.0 (2024-04-24) -- "Puppy Cup"
Rstudio (2024.04.0+735)
ggprism_1.0.5
ggplot2_3.5.1

In no particular order

• Clean up guide_prism_minor and make changes more explicit
• Remove ggplot-internal.R and replace with pulls from ggplot2 NAMESPACE (see teunbrand/ggh4x)
• Decide if annotation_ticks is useful or can be removed
• Fix tick length determination/defaults in annotation_ticks
• Decide if imports can stay in ggprism-package.R or should be moved to the functions in which they are used
• Check that scale_fill_prism has the correct structure
• Check that scale_shape_prism has the correct structure
• Make line elements in theme_prism inherit from base_line_size? Not sure about this one...
• Make axes automatically disappear if frame = TRUE in theme_prism, or at least play around with panel.border
• Add Makefile for package build
• Improve Makefile to make life easier
• Add scale_colour/color_prism functions
• Add the colour palettes that underlie scale_colour_prism
• Make file names in data-raw more consistent
• Put schemes all in the same directory
• Make README more informative
• Recreate a figure from a paper in the README
• Add .pdf with all colour schemes
• Add useful packages to suggests e.g. rstatix, ggbeeswarm, etc.
• Recreate some common figure types in a vignette

Document and add examples for:

• annotation_ticks
• ggprism-package
• ggprism_data
• guide_prism_minor
• scale_fill_prism
• scale_colour_prism
• scale_shape_prism
• theme_prism

Add tests for:

• annotation_ticks
• guide_prism_minor
• scale_fill_prism
• scale_colour_prism
• scale_shape_prism
• theme_prism

Hello

I have been using your wonderful package, unfortunately, I started getting errors across all of my Rmarkdown and scripts. This started happening about a week ago.

For example:

p_HR_all <- final_result %>% ggplot(aes(x = Time, y = HR, color=Group)) + geom_smooth(method = "loess", se = TRUE) + labs(title = "", x = "Time (seconds)", y = "HR (bpm)") + theme_prism() + scale_x_continuous(limits = c(0, 180), breaks = seq(0, 180, 60)) p_HR_all

and the error would be:

Error in plot_theme():
! The legend.text.align theme element is not defined in the element hierarchy.
Backtrace:

1. base (local) <fn>(x)
2. ggplot2:::print.ggplot(x)
3. ggplot2:::ggplot_gtable.ggplot_built(data)
4. ggplot2:::plot_theme(plot)
   Error in plot_theme(plot) :

hopefully is a quick fix. Thank you

Hello,

I'm new to plotting p-values on interaction plots, and I've been working off the piece of code below (source).

I've been trying to adjust the x-position of the brackets, such that for each pairwise comparison, the left end of the bracket is aligned with the middle of the first boxplot being considered, and the right end of the bracket is aligned with the middle of the second boxplot being considered.

In other words, I would like xmin in df_p_val to reflect the dose & supp combination of the first group in the pairwise comparison -- not just its dose.

Thank you very much in advance for any guidance you may be able to provide!

df_p_val <- ToothGrowth %>% 
  rstatix::group_by(supp) %>% 
  rstatix::t_test(len ~ dose) %>% 
  rstatix::add_xy_position()

p <- ggplot(ToothGrowth, aes(x = factor(dose), y = len)) + 
  geom_boxplot(aes(fill = supp)) + 
  theme_prism()

# remember colour and step.group.by are referring to a column name in df_p_val
p + add_pvalue(df_p_val, 
               label = "p = {p.adj}",
               colour = "supp",
               fontface = "bold",
               step.group.by = "supp",
               step.increase = 0.1,
               tip.length = 0,
               bracket.colour = "black",
               show.legend = FALSE)

add_pvalue function affects scale_y_log10's behavior, e.g.

tg <- ToothGrowth
tg$dose <- as.factor(tg$dose)
tg$group <- factor(rep(c("grp1", "grp2"), 30))
two.means <- tibble::tribble(
~group1, ~group2, ~p, ~y.position,
"OJ", "VC", 0.0606, 36)

Without scale_y_log10, add_pvalue works as expected
ggplot(tg, aes(x = supp, y = len)) + geom_boxplot() + add_pvalue(two.means)

With scale_y_log10, y-axis become extremely large and original plots are squeezed to the bottom
ggplot(tg, aes(x = supp, y = len)) + geom_boxplot() + add_pvalue(two.means) + scale_y_log10()