FAQ

Plotting FAQ

This article contains answers to common questions about plots and customizing the ggplot output of functions like plot_raw_current_data().

Note: Above all, if you are confused about how a function works, be sure to check the help page for that function! If you type ? and the name of the function in the R console, it will bring up the help page for the function. For example, type ?plot_AP_trace to get information on how to use this function.

There are often extra arguments not included in the examples that you may find helpful if you need to re-adjust parameters like the image position in plot_summary_current_data() or the bracket line width.

This is the plot we will use throughout the FAQ:

cars_plot <- ggplot(cars, aes(x = speed, y = dist)) +
  geom_point() +
  labs(x = "Speed (mph)", y = "Distance (ft)") +
  patchclampplotteR_theme()

cars_plot

How do I save my plots outside of R?

Most plots have an argument called save_png, which you can set to “yes” or “no” (the default is “no” to cut down on run time). If save_png is set to “yes”, this will save the plot as a .png using ggsave() and export it to a folder. The subfolder will vary depending on the plot type. For example, all plots generated using plot_PPR_data_one_treatment() will be exported to “Figures/Evoked-currents/PPR” relative to your project directory. If you want further control over the export options, you can also use ggsave() to manually save a ggplot object.

ggsave(
  cars_plot,
  path = here("Figures/Raw-plots"),
  file = "cars-plot.png",
  width = 7,
  height = 5,
  units = "in",
  dpi = 300
)

How do I change the plot text sizes?

A quick fix for making the text bigger is the large_axis_text argument. Set large_axis_text to “yes” to add a ggplot theme layer which increases and adjusts the plot to be ideal for posters or presentations. If you want more control over the font sizes, you can also apply a new theme() layer and specify the text elements manually. For example:

cars_plot +
  theme(text = element_text(size = 25, color = "darkmagenta"))

How do I change the plot font family?

This can be tricky and it will vary depending on your operating system. I would recommend using the extrafont package. You will need to run the code below the first time you do this. Use fonts() to see the list of fonts available to you in R.

Warning! font_import() will take a long time to run, especially if you have a lot of fonts on your computer. Luckily, you only need to do this once per system, or after you install a new font.

If the extrafont package does not work, there are newer alternatives, such as the showtext package, which makes it particularly easy to add Google fonts and system fonts.

library(extrafont)
font_import()

fonts()

If you already have the fonts installed in R, you just need to run library(extrafont) at the start of your document to have the fonts available to use.

cars_plot +
  theme(text = element_text(size = 25, color = "darkmagenta", family = "Calibri"))

To apply these theme changes globally (i.e. all plots in your file), it will be much more efficient to modify and define a theme. See below!

How do I modify the ggplot theme?

You might wonder things like, “How can I override the default theme that came with this package?” or “How can I make all my plots have the same theme/fonts, etc.?”. You can fix this efficiently by defining a custom ggplot theme and adding this to the ggplot_theme argument in all plotting functions.

Choose a theme like theme_classic() or the patchclampplotteR_theme() that came with this package. Use this theme as a base, and add new theme elements to replace just the components you want to change (see the documentation on theme modification in ggplot2. The new theme that you define will inherit all components of the base theme, but replace just the elements that you specified.

Important! You must assign this theme to a name, since the ggplot_theme argument requires a named object. In this example, I will creatively call it my_new_theme.

my_new_theme <- patchclampplotteR_theme() %+replace%
  theme(
    text = element_text(size = 25, color = "#333333"),
    axis.title = element_text(face = "plain", size = 14),
    panel.grid.major = element_line(color = "#ebebeb"),
    axis.text = element_text(face = "italic", size = 8, color = "#999999"),
    axis.line = element_blank()
  )

cars_plot + my_new_theme

Now, insert this theme into the ggplot_theme argument. This is one of the sample action potential plots with the default theme applied:

plot_AP_comparison(
  sample_AP_data,
  plot_treatment = "Control",
  plot_category = 2,
  y_variable = "peak_amplitude",
  y_axis_title = "Peak Amplitude (pA)",
  theme_options = sample_theme_options,
  baseline_label = "Baseline",
  test_type = "wilcox.test",
  post_hormone_label = "Insulin",
  treatment_colour_theme = sample_treatment_names_and_colours,
  save_plot_png = "no",
  ggplot_theme = patchclampplotteR_theme()
)

Here’s what the same plot looks like with the new theme:

plot_AP_comparison(
  sample_AP_data,
  plot_treatment = "Control",
  plot_category = 2,
  y_variable = "peak_amplitude",
  y_axis_title = "Peak Amplitude (pA)",
  theme_options = sample_theme_options,
  baseline_label = "Baseline",
  test_type = "wilcox.test",
  post_hormone_label = "Insulin",
  treatment_colour_theme = sample_treatment_names_and_colours,
  save_plot_png = "no",
  ggplot_theme = my_new_theme
)

How do I show one sex only?

You can use the included_sexes argument in plot_summary_current_data(). It defaults to “both”, but if you set it to “male” or “female”, the data will automatically be filtered to show one sex only, and the resulting .png will have “-males-only” or “-females-only” included in the filename.

How can I plot many treatments at once and/or plot even faster?

Let’s say that you are running a plotting function repeatedly, but only changing one argument at once (like plot_treatment). To reduce repetitive code, you can use pmap() from the purrr package to run a function with a list of treatments.

First, I am adding a new column to sample_treatment_names_and_colours which contains the filepath to each representative trace .png file.

evoked_trace_filename <- c(
  "vignettes/articles/figures/Category-2-Control-Trace.png",
  "vignettes/articles/figures/Category-2-HNMPA-Trace.png",
  "vignettes/articles/figures/Category-2-PPP-Trace.png",
  "vignettes/articles/figures/Category-2-PPP-and-HNMPA-Trace.png"
)

list_of_treatments <- sample_treatment_names_and_colours

list_of_treatments$representative_trace_filepath <- evoked_trace_filename

Then, I’m using purrr::pmap() and inserting the list_of_treatments that I defined above. This is so I can use the treatment column in the plot_treatment argument and the representative_trace_filepath column for the representative_trace_filename argument. See below - the summary plots for all treatments were produced with one code chunk!

pmap(
  .l = list_of_treatments,
  .f = ~ with(
    list(...),
    plot_summary_current_data(
      data = sample_pruned_eEPSC_df$all_cells,
      plot_category = 2,
      plot_treatment = treatment,
      current_type = "eEPSC",
      y_variable = "amplitude",
      hormone_added = "Insulin",
      hormone_or_HFS_start_time = 5,
      include_representative_trace = "yes",
      representative_trace_filename = representative_trace_filepath,
      y_axis_limit = 175,
      signif_stars = "yes",
      t_test_df = sample_eEPSC_t_test_df,
      large_axis_text = "no",
      shade_intervals = "no",
      treatment_colour_theme = sample_treatment_names_and_colours,
      theme_options = sample_theme_options
    )
  )
)
#> [[1]]

#> 
#> [[2]]

#> 
#> [[3]]

#> 
#> [[4]]

How do I only show some treatments?

Functions with multiple treatments (such as plot_PPR_data_multiple_treatments() and plot_AP_frequencies_multiple_treatments()) have an argument called include_all_treatments. Change this to “no” and then define a list of treatments in the list_of_treatments argument.

plot_AP_frequencies_multiple_treatments(
  data = sample_AP_count_data,
  include_all_treatments = "no",
  list_of_treatments = c("PPP", "HNMPA"),
  plot_category = 2,
  treatment_colour_theme = sample_treatment_names_and_colours
)
#> `geom_smooth()` using method = 'loess' and formula = 'y ~ x'

Theme FAQ

How do I insert my own colours and treatment names?

You will define your treatment names and colours once at the start, then refer to it using the treatment_colour_theme in all plotting functions.

This package comes pre-loaded with sample_treatment_names_and_colours:

First, check out how many treatment groups you have using unique(raw_eEPSC_df$treatment).

unique(sample_raw_eEPSC_df$treatment)
#> [1] Control       HNMPA         PPP           PPP_and_HNMPA
#> Levels: Control HNMPA PPP PPP_and_HNMPA

Next, modify this code to set up your own dataframe with your own hormone names and colours.

my_theme_colours <- data.frame(
  treatment = c("Control", "HNMPA", "PPP", "PPP_and_HNMPA"),
  display_names = c("Control", "HNMPA", "PPP", "PPP\n&\nHNMPA"),
  colours = c("#f07e05", "#f50599", "#70008c", "#DCE319"),
  very_pale_colours = c("#fabb78", "#fa98d5", "#ce90de", "yellow")
)

Every time a plot contains the argument treatment_colour_theme, refer to your custom dataframe.

plot_PPR_data_multiple_treatments(
  data = sample_PPR_df,
  include_all_treatments = "yes",
  plot_category = 2,
  baseline_label = "B",
  post_hormone_label = "I",
  test_type = "t.test",
  theme_options = sample_theme_options,
  treatment_colour_theme = my_theme_colours
)

How do I modify the theme_options?

This dataset contains sample_theme_options that will affect all plots. These include values for line thickness and the point shapes for male vs. female data points:

To use the default values, just use sample_theme_options in the theme_options argument of any plotting function in patchclampplotteR.

plot_summary_current_data(
  data = sample_pruned_eEPSC_df$all_cells,
  plot_category = 2,
  plot_treatment = "Control",
  current_type = "eEPSC",
  y_variable = "amplitude",
  hormone_added = "Insulin",
  hormone_or_HFS_start_time = 5,
  include_representative_trace = "no",
  y_axis_limit = 175,
  signif_stars = "yes",
  t_test_df = sample_eEPSC_t_test_df,
  large_axis_text = "no",
  shade_intervals = "no",
  treatment_colour_theme = sample_treatment_names_and_colours,
  theme_options = sample_theme_options
)
#> Warning: Removed 25 rows containing missing values or values outside the scale range
#> (`geom_segment()`).

If you want to change these values, please follow these steps.

Step 1: Create a .csv file with two columns (option and value), modelled after this sample dataset.

Important!: Your .csv must have identical columns and rows as the sample data, or else the plots won’t work!

Step 2: Read in the .csv file in using utils::read_csv(). This will now be an object in your R environment.

Step 3: Important!! You must convert the first column (option) into the rownames. This is a mandatory step to allow the theme_options to be indexed by row name in plotting functions.

Step 4: Run the following code:

library(tibble)

my_custom_theme_options <- read.csv("my_custom_theme_options.csv") %>%
  remove_rownames() %>%
  column_to_rownames(var = "option")

Step 5: Check the resulting object. You should now have 11 objects of 1 variable, and the row names should be gray_shading_colour, line_col, etc.

Step 6: Go to a plotting function like plot_raw_current_data() and replace sample_theme_options with your newly created object from Step 5.

I changed line_col to orange, and I also changed male_shape and female_shape. See, the graph looks different now!

plot_summary_current_data(
  data = sample_pruned_eEPSC_df$all_cells,
  plot_category = 2,
  plot_treatment = "Control",
  current_type = "eEPSC",
  y_variable = "amplitude",
  hormone_added = "Insulin",
  hormone_or_HFS_start_time = 5,
  include_representative_trace = "no",
  y_axis_limit = 175,
  signif_stars = "yes",
  t_test_df = sample_eEPSC_t_test_df,
  large_axis_text = "no",
  shade_intervals = "no",
  treatment_colour_theme = sample_treatment_names_and_colours,
  theme_options = my_custom_theme_options
)
#> Warning: Removed 25 rows containing missing values or values outside the scale range
#> (`geom_segment()`).

What does this package replace?

This package replaces manual plot creation and data analysis in GraphPad Prism and/or Excel. Previously, data analysis would have required transferring data between programs, which is time-consuming and potentially error-prone. This package can be combined with report writing in R to generate fully reproducible manuscripts.

Package FAQ

Who’s Ruby?

Ruby is a rat, and she’s also become an unofficial lab mascot for the Crosby lab! She loves helping out with data analysis, welcoming newcomers to the lab, and keeping lab members company. She found a way to sneak into some of the documentation for this package!

Ruby says ‘Hi humans!!’