├── .gitignore
├── README.md
├── ggplot-tutorial.Rmd
├── ggplot-tutorial.html
└── ggplot2-for-publications.Rproj
/.gitignore:
--------------------------------------------------------------------------------
1 | .Rproj.user
2 | .Rhistory
3 | .RData
4 | .Ruserdata
5 |
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # Making figures ready for publication with `ggplot2`
2 |
3 | This tutorial offers a step-by-step guide for how to create publication-ready figures using `ggplot2` and the data from `palmerpenguins`.
4 |
5 | 
6 |
7 | ## Install the package & data
8 |
9 | ```{r, message = FALSE}
10 | # Install the package
11 | remotes::install_github("allisonhorst/palmerpenguins")
12 |
13 | # Load the package
14 | library(palmerpenguins)
15 |
16 | # Load the data into the Global Environment
17 | data("penguins")
18 | ```
19 |
20 | ## Meet the Penguins
21 |
22 | 
23 |
24 | Artwork by @allison_horst
25 |
26 |
27 | ## What are culmen length & depth?
28 |
29 | The culmen is the upper ridge of a bird’s bill. In the simplified penguins data, culmen length and depth are renamed as variables bill_length_mm and bill_depth_mm to be more intuitive.
30 | For this penguin data, the culmen (bill) length and depth are measured as shown below:
31 |
32 | 
33 |
34 | Artwork by @allison_horst
35 |
36 |
37 | ## The steps for creating a beautiful scatter plot in `ggplot2`
38 |
39 | First we will create a basic scatterplot of `body_mass_g` against `bill_length_mm`.
40 |
41 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
42 | # Load the package
43 | library(ggplot2)
44 |
45 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+ # this is the data
46 | geom_point() # here we add the points
47 | ```
48 |
49 | 
50 |
51 | ### Change the size of points
52 |
53 | We can manually change the size of our datapoints. The points in the standard plot are quite small, so lets increase the size of the points with `size = 3`.
54 |
55 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
56 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
57 | geom_point(size = 3)
58 | ```
59 |
60 | 
61 |
62 |
63 | ### Change the shape of points
64 |
65 | In `ggplot2`, it is possible to change the shape of the points. Here is a quick reference guide:
66 |
67 | 
68 |
69 | The shape of all datapoints can be changed with e.g. `shape = 8`.
70 |
71 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
72 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
73 | geom_point(size = 3, shape = 8)
74 | ```
75 |
76 | 
77 |
78 |
79 | Alternatively, we can change the shape of our points based on species with `aes(shape = species)`.
80 |
81 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
82 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
83 | geom_point(aes(shape = species), size = 3)
84 | ```
85 |
86 | 
87 |
88 | ### Change the opacity of points
89 |
90 | You can also change the opacity of the data points using `alpha`. Alpha values are required to be between 0 - 1 where 0 is transparent and 1 is opaque.
91 |
92 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
93 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
94 | geom_point(aes(shape = species), size = 3, alpha = 0.6)
95 | ```
96 |
97 | 
98 |
99 | ### Adding colour
100 |
101 | Now lets explore the different species by adding colour with the code `colour = species`.
102 |
103 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
104 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
105 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)
106 | ```
107 |
108 | 
109 |
110 | This red-green colour combination is colourblind unfrieldly, so lets change the colour of the points with `scale_colour_manual`. To ensure the shapes match with the names we will also use `scale_shape_manual`.
111 |
112 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
113 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
114 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)+
115 |
116 | scale_colour_manual(values = c("#C15CCB", "#00868B", "#FF6A00"),
117 | labels = c("Chinstrap", "Gentoo", "Adélie"))+
118 | scale_shape_manual(values = c(17, 15, 16),
119 | labels = c("Chinstrap", "Gentoo", "Adélie"))
120 | ```
121 |
122 | 
123 |
124 | We won't change the points any more, so let's save the plot as `penguin_plot`, so we can build upon it.
125 |
126 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
127 | penguin_plot <- ggplot(penguins, aes(body_mass_g, bill_length_mm))+
128 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)+
129 |
130 | scale_colour_manual(values = c("#C15CCB", "#00868B", "#FF6A00"),
131 | labels = c("Chinstrap", "Gentoo", "Adélie"))+
132 | scale_shape_manual(values = c(17, 15, 16),
133 | labels = c("Chinstrap", "Gentoo", "Adélie"))
134 | ```
135 |
136 | ### Changing the background
137 |
138 | You can change the background of `ggplot2` figures in a variety of ways with:
139 |
140 | - `theme_gray()`
141 | - `theme_bw()`
142 | - `theme_linedraw()`
143 | - `theme_light()`
144 | - `theme_minimal()`
145 | - `theme_classic()`
146 | - `theme_void()`
147 | - `theme_dark()`
148 |
149 | 
150 |
151 |
152 | My personal favourite is `theme_bw` therefore we will continue to make our plot with this theme.
153 |
154 | We will further remove the thicker lines in the background with `panel.grid.major = element_blank()`, and the thinner lines with `panel.grid.minor = element_blank()`.
155 |
156 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
157 | penguin_plot +
158 | theme_bw()+ # set the background theme
159 | theme(panel.grid.major = element_blank(), # remove the major lines
160 | panel.grid.minor = element_blank()) # remove the minor lines
161 | ```
162 |
163 | 
164 |
165 | ### Changing the text size
166 |
167 | There are several ways to change the size of the font, but we can quickly change all font size with `theme_bw(base_size = 20)`.
168 |
169 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
170 | penguin_plot <- penguin_plot +
171 | theme_bw(base_size = 15)+
172 | theme(panel.grid.major = element_blank(),
173 | panel.grid.minor = element_blank())
174 |
175 | penguin_plot
176 | ```
177 |
178 | 
179 |
180 | ### Renaming the axes and legend
181 |
182 | You can rename the axes and legend using `labs`.
183 |
184 |
185 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
186 | penguin_plot <- penguin_plot +
187 | labs(x = "Body Mass (g)", y = "Bill Length (mm)", colour = "Species", shape = "Species")
188 |
189 | penguin_plot
190 | ```
191 |
192 |
193 | 
194 |
195 | ### Change axes position
196 |
197 | In the standard plots, the axes titles are really close to the plot. We will increase their distance with `vjust`:
198 |
199 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
200 | penguin_plot <- penguin_plot +
201 | theme(axis.title.y = element_text(vjust = 3))+ # increase distance from the y-axis
202 | theme(axis.title.x = element_text(vjust = -1)) # increase distance from the x-axis
203 |
204 | penguin_plot
205 | ```
206 |
207 | 
208 |
209 | ### Change legend position
210 |
211 | Legends can be positioned in a number of different ways using `legend.position`:
212 |
213 | - `theme(legend.position="top")`
214 | - `theme(legend.position="bottom")`
215 | - `theme(legend.position="left")`
216 | - `theme(legend.position="right")`
217 | - `theme(legend.position="none")`
218 |
219 | Legend loction can be also a numeric vector c(x,y), where x and y are the coordinates of the legend box. Their values should be between 0 - 1. c(0,0) is the “bottom left” and c(1,1) is the “top right” position.
220 |
221 |
222 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
223 | penguin_plot <- penguin_plot +
224 | theme(legend.position = c(0.85, 0.2))
225 |
226 | penguin_plot
227 | ```
228 |
229 | 
230 |
231 | ### Saving the figure
232 |
233 | Publications require high quality images and they specify the size and format required on their website.
234 |
235 | #### `ggsave`
236 | `ggsave` is one example of how to save your figures.
237 |
238 | It allows you to save high quality images in a variety of different file types (e.g. "png", "eps", "ps", "tex", "pdf", "jpeg", "tiff", "png", "bmp", "svg", "wmf"). You can also specify the `width` and `height` in "in", "cm", or "mm", and specify the plot resolution with `dpi`.
239 |
240 | ```{r, eval = FALSE}
241 | # This will save the last plot for the code you ran:
242 |
243 | ggsave("penguin_plot.pdf",
244 | dpi = 600,
245 | width = 100, height = 60, unit = "mm")
246 | ```
247 |
248 | #### `pdf`
249 |
250 | An alternative method is to use `pdf`. This allows you to specify the colour mode (e.g. cmyk).
251 |
252 | ```{r, eval = FALSE}
253 | pdf("ggplot-cmyk.pdf", width = 12 / 2.54, height = 8 / 2.54,
254 | colormodel = "cmyk")
255 |
256 | print(penguin_plot)
257 |
258 | dev.off()
259 | ```
260 |
261 |
262 |
263 | ## More `ggplot2` plots and tricks
264 |
265 |
266 | ### Change the size of points based on other data
267 |
268 | Size can be changed based on data within the `penguins` dataframe. For example, here we have changed the size of the points based on `bill_depth_mm`.
269 |
270 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
271 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
272 | geom_point(aes(size = bill_depth_mm))
273 | ```
274 | 
275 |
276 | ### Adding an ellispe
277 |
278 | Add 95% confidence interval ellispses with `stat_ellipse`.
279 |
280 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
281 | ellipse<- penguin_plot +
282 | stat_ellipse(aes(colour = species, level=0.95))
283 |
284 |
285 | ellipse
286 | ```
287 |
288 | 
289 |
290 |
291 | ### Linear regression line
292 |
293 | Add a linear regression line with `geom_smooth(method=lm)`.
294 |
295 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
296 | lm <- penguin_plot +
297 | geom_smooth(method=lm, aes(colour = species))
298 |
299 | lm
300 | ```
301 | 
302 |
303 | ### `facet_wrap`
304 |
305 | `facet_wrap` is a fantastic tool to slit plots based on a specified categorical column. Here we will use the `species` column.
306 |
307 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
308 | penguin_plot +
309 | facet_wrap(~species, ncol = 3, nrow = 1) + # specifying 3 columns, 1 row
310 | theme(legend.position = "none") # remove legend
311 | ```
312 |
313 | 
314 |
315 |
316 | ### Changing strip design
317 |
318 | It is also possible to change the `size`, `colour`, `face` and `fill` colour of the facet strips with the following code:
319 |
320 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
321 | penguin_plot +
322 | facet_wrap(~species, ncol = 3, nrow = 1) + # specifying 3 columns, 1 row
323 | theme(legend.position = "none")+ # remove legend
324 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
325 | strip.background = element_rect(fill="black"))
326 | ```
327 |
328 | 
329 |
330 |
331 | ### Removing white space and free axes
332 |
333 | We can also remove the free space using `scales = free`.
334 |
335 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
336 | penguin_plot +
337 | facet_wrap(~species, scales = "free")+
338 | theme(legend.position = "none")+ # remove legend
339 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
340 | strip.background = element_rect(fill="black"))
341 | ```
342 |
343 | 
344 |
345 |
346 | ### `facet_grid`: Free facet width
347 |
348 | To allow the facets to be different widths, we must use `facet_grid` and `space = "free"`. This can be used with `scales = "free"`.
349 |
350 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
351 | penguin_plot +
352 | facet_grid(.~species, scales = "free", space = "free") +
353 | theme(legend.position = "none")+ # remove legend
354 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
355 | strip.background = element_rect(fill="black"))
356 | ```
357 | 
358 |
359 | ### Bar plot
360 |
361 |
362 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
363 | library(dplyr)
364 | penguin_summary <- penguins %>%
365 | group_by(species) %>%
366 | summarise(mean = mean(body_mass_g, na.rm = T),
367 | sd = sd(body_mass_g, na.rm = T))
368 |
369 |
370 | ggplot(penguin_summary, aes(y = mean, x=species, fill = species)) +
371 | geom_bar(stat="identity")+
372 | geom_errorbar(aes(ymin = mean-sd, ymax = mean+sd),
373 | width=.1)
374 | ```
375 | 
376 |
377 | Create the plot like the code above for the scatter plot.
378 |
379 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
380 | bar <- ggplot(penguin_summary, aes(y = mean, x = species, fill = species)) +
381 | geom_bar(stat = "identity")+
382 | geom_errorbar(aes(ymin = mean-sd, ymax = mean+sd),
383 | width=.1)+
384 | theme_bw(base_size = 20)+
385 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
386 | labs(x = "Species", y = "Body Mass (g)")+
387 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
388 | theme(legend.position = "none") +
389 | scale_y_continuous(limits = c(0, 5700))+
390 | theme(axis.title.y = element_text(vjust = 3)) +
391 | theme(axis.title.x = element_text(vjust = -1))
392 |
393 | bar
394 | ```
395 | 
396 |
397 | ### Histogram
398 |
399 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
400 | ggplot(penguins, aes(x = flipper_length_mm, fill = species)) +
401 | geom_histogram(alpha = 0.4)+
402 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))
403 | ```
404 |
405 | 
406 |
407 | ### Density Plot
408 |
409 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
410 | ggplot(penguins, aes(x = flipper_length_mm, fill = species)) +
411 | geom_density(alpha = 0.4)+
412 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))
413 | ```
414 |
415 | 
416 |
417 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
418 | flipper <- ggplot(penguins, aes(x = flipper_length_mm, fill = species, colour = species)) +
419 | geom_density(alpha = 0.4)+
420 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
421 | scale_colour_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
422 | theme_bw(base_size = 20)+
423 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
424 | labs(x = "Flipper Length (mm)", y = "Density", fill = "Species", colour = "Species")+
425 | #theme(legend.position = "none") +
426 | scale_x_continuous(expand = c(0,0), limits = c(165, 238))+
427 | scale_y_continuous(expand = c(0,0), limits = c(0, 0.065), breaks=seq(0, 0.065, 0.02))+
428 | theme(axis.title.y = element_text(vjust = 3)) +
429 | theme(axis.title.x = element_text(vjust = -1))
430 |
431 | flipper
432 | ```
433 |
434 | 
435 |
436 | ### Add a mean line
437 |
438 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
439 | flipper_mean <- penguins %>%
440 | group_by(species) %>%
441 | summarise(mean = mean(flipper_length_mm, na.rm = TRUE))
442 |
443 | density <- flipper +
444 | geom_vline(data = flipper_mean, aes(xintercept = mean, colour = species), linetype = "dashed", size = 1.3)
445 |
446 | density
447 | ```
448 |
449 | 
450 |
451 |
452 | ### Box plot
453 |
454 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
455 | ggplot(na.omit(penguins), aes(x=species, y=flipper_length_mm, fill=sex)) +
456 | geom_boxplot()
457 | ```
458 | 
459 |
460 |
461 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
462 | box_plot<- ggplot(na.omit(penguins), aes(x=species, y=flipper_length_mm, fill=sex)) +
463 | geom_boxplot()+
464 | theme_bw(base_size = 20)+
465 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
466 | labs(x = "Species", y = "Flipper Length (mm)", fill = "Sex")+
467 | theme(axis.title.y = element_text(vjust = 3)) +
468 | theme(axis.title.x = element_text(vjust = -1)) +
469 | theme(legend.position = c(0.15, 0.8))
470 |
471 | box_plot
472 | ```
473 | 
474 |
475 |
476 | ### Arrange plots
477 |
478 | There are several different methods that allow you to arrange your plots into different panels.
479 |
480 | #### `ggarrange`
481 |
482 | Here we will arrange four of the plots that we made above into one figure using `ggarrange`. You can specify the number of rows `nrow` and `ncol` and add `labels.
483 |
484 | Note we previously saved our figures (`ellipse`, `density`, `bar`, `box_plot`) in the Global Environment and we are calling on them here.
485 |
486 |
487 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=12, fig.height=8}
488 | library(ggpubr)
489 |
490 | ggarrange(ellipse, density, bar, box_plot,
491 | nrow = 2, ncol = 2,
492 | labels = c("a", "b", "c", "d"))
493 | ```
494 |
495 |
496 | 
497 |
498 |
499 | #### Same axes
500 |
501 | If two plots have the same axes, you can remove one and use `align = "v"` to ensure the plots remain the same size.
502 |
503 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=12, fig.height=8}
504 | # Remove y axes text and title
505 | lm <- lm +
506 | theme(axis.title.y=element_blank(),
507 | axis.text.y=element_blank(),
508 | axis.ticks.y=element_blank())
509 |
510 |
511 | ggarrange(ellipse, lm,
512 | nrow = 1, ncol = 2,
513 | labels = c("a", "b"),
514 | align = "v")
515 | ```
516 |
517 | 
518 |
519 |
520 | #### `patchwork`
521 |
522 | `patchwork` is a very straighforward and intuative package for arranging plots.
523 | e.g:
524 |
525 | - `plot1 + plot2` aligns two plots next to each other
526 | - `plot1 / plot2` aligns plot2 under plot1
527 |
528 | Lets make a fancy one.
529 |
530 | We add the figure labels with e.g. `labs(tag = 'a')`. We will also make two of our plots half the size of the main plot with `plot_layout(widths=c(2,1))`.
531 |
532 |
533 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=18, fig.height=10}
534 | library(patchwork)
535 |
536 | (ellipse + labs(tag = 'a')| ((bar+ labs(tag = 'b')) / (box_plot +labs(tag = 'c')))) + plot_layout(widths=c(2,1))
537 | ```
538 | 
539 |
540 |
541 | 
542 |
543 | Artwork by @CerrenRichards
544 |
545 |
--------------------------------------------------------------------------------
/ggplot-tutorial.Rmd:
--------------------------------------------------------------------------------
1 | ---
2 | title: "ggplot2-tutorial"
3 | author: "Cerren Richards"
4 | date: "7/4/2020"
5 | output: html_document
6 | ---
7 |
8 | ```{r setup, include=FALSE}
9 | knitr::opts_chunk$set(echo = TRUE)
10 | ```
11 |
12 | ## Overview
13 |
14 | This tutorial offers a step-by-step guide for how to create publication-ready figures using `ggplot2` and the data from `palmerpenguins`.
15 |
16 |
17 | ## Install the package & data
18 |
19 | ```{r, message = FALSE}
20 |
21 | # Install the package
22 | remotes::install_github("allisonhorst/palmerpenguins")
23 |
24 | # Load the package
25 | library(palmerpenguins)
26 |
27 | # Load the data into the Global Environment
28 | data("penguins")
29 |
30 | # View the data
31 | head(penguins)
32 |
33 | ```
34 |
35 |
36 |
37 | ## The steps for creating a beautiful scatter plot in `ggplot2`
38 |
39 | First we will create a basic scatterplot of `body_mass_g` against `bill_length_mm`.
40 |
41 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
42 |
43 | # Load the package
44 | library(ggplot2)
45 |
46 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+ # this is the data
47 | geom_point() # here we add the points
48 |
49 | ```
50 |
51 |
52 | ### Change the size of points
53 |
54 | We can manually change the size of our datapoints. The points in the standard plot are quite small, so lets increase the size of the points with `size = 3`.
55 |
56 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
57 |
58 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
59 | geom_point(size = 3)
60 |
61 | ```
62 |
63 |
64 |
65 | ### Change the shape of points
66 |
67 | In `ggplot2`, it is possible to change the shape of the points. Here is a quick reference guide:
68 |
69 | ```{r, echo=FALSE, fig.align = "center"}
70 | #include_graphics("shapes.png")
71 | ```
72 |
73 | The shape of all datapoints can be changed with e.g. `shape = 8`.
74 |
75 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
76 |
77 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
78 | geom_point(size = 3, shape = 8)
79 |
80 | ```
81 |
82 |
83 | Alternatively, we can change the shape of our points based on species with `aes(shape = species)`.
84 |
85 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
86 |
87 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
88 | geom_point(aes(shape = species), size = 3)
89 |
90 | ```
91 |
92 |
93 | ### Change the opacity of points
94 |
95 | You can also change the opacity of the data points using `alpha`. Alpha values are required to be between 0 - 1 where 0 is transparent and 1 is opaque.
96 |
97 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
98 |
99 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
100 | geom_point(aes(shape = species), size = 3, alpha = 0.6)
101 |
102 | ```
103 |
104 |
105 |
106 | ### Adding colour
107 |
108 | Now lets explore the different species by adding colour with the code `colour = species`.
109 |
110 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
111 |
112 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
113 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)
114 |
115 |
116 | ```
117 |
118 |
119 | This red-green colour combination is colourblind unfrieldly, so lets change the colour of the points with `scale_colour_manual`. To ensure the shapes match with the names we will also use `scale_shape_manual`.
120 |
121 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
122 |
123 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
124 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)+
125 |
126 | scale_colour_manual(values = c("#C15CCB", "#00868B", "#FF6A00"),
127 | labels = c("Chinstrap", "Gentoo", "Adélie"))+
128 | scale_shape_manual(values = c(17, 15, 16),
129 | labels = c("Chinstrap", "Gentoo", "Adélie"))
130 |
131 |
132 | ```
133 |
134 |
135 | We won't change the points any more, so let's save the plot as `penguin_plot`, so we can build upon it.
136 |
137 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
138 |
139 | penguin_plot <- ggplot(penguins, aes(body_mass_g, bill_length_mm))+
140 | geom_point(aes(shape = species, colour = species), size = 3, alpha = 0.6)+
141 |
142 | scale_colour_manual(values = c("#C15CCB", "#00868B", "#FF6A00"),
143 | labels = c("Chinstrap", "Gentoo", "Adélie"))+
144 | scale_shape_manual(values = c(17, 15, 16),
145 | labels = c("Chinstrap", "Gentoo", "Adélie"))
146 |
147 | ```
148 |
149 | ### Changing the background
150 |
151 | You can change the background of `ggplot2` figures in a variety of ways with:
152 |
153 | - `theme_gray()`
154 | - `theme_bw()`
155 | - `theme_linedraw()`
156 | - `theme_light()`
157 | - `theme_minimal()`
158 | - `theme_classic()`
159 | - `theme_void()`
160 | - `theme_dark()`
161 |
162 |
163 | ```{r, echo=FALSE, warning=FALSE, message = FALSE, fig.align = "center"}
164 |
165 |
166 | plot <- ggplot(penguins, aes(body_mass_g, bill_length_mm))+
167 | geom_point( alpha = 0.3)
168 |
169 | a <- plot + theme_gray() + ggtitle("theme_gray")
170 | b <- plot + theme_bw() + ggtitle("theme_bw")
171 | c <- plot + theme_linedraw() + ggtitle("theme_linedraw")
172 | d <- plot + theme_light() + ggtitle("theme_light")
173 | e <- plot + theme_minimal() + ggtitle("theme_minimal")
174 | f <- plot + theme_classic() + ggtitle("theme_classic")
175 | g <- plot + theme_void() + ggtitle("theme_void")
176 | h <- plot + theme_dark() + ggtitle("theme_dark")
177 |
178 | library(ggpubr)
179 |
180 | ggarrange(a, b, c, d,
181 | nrow = 2, ncol = 2)
182 |
183 | ggarrange( e, f, g, h,
184 | nrow = 2, ncol = 2)
185 |
186 | ```
187 |
188 |
189 |
190 | My personal favourite is `theme_bw` therefore we will continue to make our plot with this theme.
191 |
192 | We will further remove the thicker lines in the background with `panel.grid.major = element_blank()`, and the thinner lines with `panel.grid.minor = element_blank()`.
193 |
194 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
195 | penguin_plot +
196 | theme_bw()+ # set the background theme
197 | theme(panel.grid.major = element_blank(), # remove the major lines
198 | panel.grid.minor = element_blank()) # remove the minor lines
199 | ```
200 |
201 |
202 |
203 | ### Changing the text size
204 |
205 | There are several ways to change the size of the font, but we can quickly change all font size with `theme_bw(base_size = 20)`.
206 |
207 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
208 | penguin_plot <- penguin_plot +
209 | theme_bw(base_size = 15)+
210 | theme(panel.grid.major = element_blank(),
211 | panel.grid.minor = element_blank())
212 |
213 | penguin_plot
214 | ```
215 |
216 | ### Renaming the axes and legend
217 |
218 | You can rename the axes and legend using `labs`.
219 |
220 |
221 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
222 | penguin_plot <- penguin_plot +
223 | labs(x = "Body Mass (g)", y = "Bill Length (mm)", colour = "Species", shape = "Species")
224 |
225 | penguin_plot
226 | ```
227 |
228 | ### Change axes position
229 |
230 | In the standard plots, the axes titles are really close to the plot. We will increase their distance with `vjust`:
231 |
232 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
233 |
234 | penguin_plot <- penguin_plot +
235 | theme(axis.title.y = element_text(vjust = 3))+ # increase distance from the y-axis
236 | theme(axis.title.x = element_text(vjust = -1)) # increase distance from the x-axis
237 |
238 | penguin_plot
239 | ```
240 |
241 |
242 |
243 | ### Change legend position
244 |
245 | Legends can be positioned in a number of different ways using `legend.position`:
246 |
247 | - `theme(legend.position="top")`
248 | - `theme(legend.position="bottom")`
249 | - `theme(legend.position="left")`
250 | - `theme(legend.position="right")`
251 | - `theme(legend.position="none")`
252 |
253 | Legend loction can be also a numeric vector c(x,y), where x and y are the coordinates of the legend box. Their values should be between 0 - 1. c(0,0) is the “bottom left” and c(1,1) is the “top right” position.
254 |
255 |
256 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
257 | penguin_plot <- penguin_plot +
258 | theme(legend.position = c(0.85, 0.2))
259 |
260 | penguin_plot
261 | ```
262 |
263 |
264 | ### Saving the figure
265 |
266 | Publications require high quality images and they specify the size and format required on their website.
267 |
268 | #### `ggsave`
269 | `ggsave` is one example of how to save your figures.
270 |
271 | It allows you to save high quality images in a variety of different file types (e.g. "png", "eps", "ps", "tex", "pdf", "jpeg", "tiff", "png", "bmp", "svg", "wmf"). You can also specify the `width` and `height` in "in", "cm", or "mm", and specify the plot resolution with `dpi`.
272 |
273 | ```{r, eval = FALSE}
274 |
275 | # This will save the last plot for the code you ran:
276 |
277 | ggsave("penguin_plot.pdf",
278 | dpi = 600,
279 | width = 100, height = 60, unit = "mm")
280 |
281 | ```
282 |
283 | #### `pdf`
284 |
285 | An alternative method is to use `pdf`. This allows you to specify the colour mode (e.g. cmyk).
286 |
287 | ```{r, eval = FALSE}
288 |
289 | pdf("ggplot-cmyk.pdf", width = 12 / 2.54, height = 8 / 2.54,
290 | colormodel = "cmyk")
291 |
292 | print(penguin_plot)
293 |
294 | dev.off()
295 |
296 | ```
297 |
298 |
299 |
300 | ## More `ggplot2` plots and tricks
301 |
302 |
303 | ### Change the size of points based on other data
304 |
305 | Size can be changed based on data within the `penguins` dataframe. For example, here we have changed the size of the points based on `bill_depth_mm`.
306 |
307 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
308 |
309 | ggplot(penguins, aes(body_mass_g, bill_length_mm))+
310 | geom_point(aes(size = bill_depth_mm))
311 |
312 | ```
313 |
314 |
315 | ### Adding an ellispe
316 |
317 | Add 95% confidence interval ellispses with `stat_ellipse`.
318 |
319 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
320 |
321 | ellipse<- penguin_plot +
322 | stat_ellipse(aes(colour = species, level=0.95))
323 |
324 |
325 | ellipse
326 | ```
327 |
328 | ### Linear regression line
329 |
330 | Add a linear regression line with `geom_smooth(method=lm)`.
331 |
332 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
333 |
334 | lm <- penguin_plot +
335 | geom_smooth(method=lm, aes(colour = species))
336 |
337 | lm
338 |
339 | ```
340 |
341 |
342 | ### `facet_wrap`
343 |
344 | `facet_wrap` is a fantastic tool to slit plots based on a specified categorical column. Here we will use the `species` column.
345 |
346 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
347 |
348 | penguin_plot +
349 | facet_wrap(~species, ncol = 3, nrow = 1) + # specifying 3 columns, 1 row
350 | theme(legend.position = "none") # remove legend
351 |
352 | ```
353 |
354 | ### Changing strip design
355 |
356 | It is also possible to change the `size`, `colour`, `face` and `fill` colour of the facet strips with the following code:
357 |
358 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
359 |
360 | penguin_plot +
361 | facet_wrap(~species, ncol = 3, nrow = 1) + # specifying 3 columns, 1 row
362 | theme(legend.position = "none")+ # remove legend
363 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
364 | strip.background = element_rect(fill="black"))
365 |
366 | ```
367 |
368 | ### Removing white space and free axes
369 |
370 | We can also remove the free space using `scales = free`.
371 |
372 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
373 |
374 | penguin_plot +
375 | facet_wrap(~species, scales = "free")+
376 | theme(legend.position = "none")+ # remove legend
377 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
378 | strip.background = element_rect(fill="black"))
379 |
380 | ```
381 |
382 | ### `facet_grid`: Free facet width
383 |
384 | To allow the facets to be different widths, we must use `facet_grid` and `space = "free"`. This can be used with `scales = "free"`.
385 |
386 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
387 |
388 | penguin_plot +
389 | facet_grid(.~species, scales = "free", space = "free") +
390 | theme(legend.position = "none")+ # remove legend
391 | theme(strip.text.x = element_text(size = 16, color = "white", face = "bold"),
392 | strip.background = element_rect(fill="black"))
393 |
394 | ```
395 |
396 |
397 | ### Bar plot
398 |
399 |
400 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
401 | library(dplyr)
402 | penguin_summary <- penguins %>%
403 | group_by(species) %>%
404 | summarise(mean = mean(body_mass_g, na.rm = T),
405 | sd = sd(body_mass_g, na.rm = T))
406 |
407 |
408 | ggplot(penguin_summary, aes(y = mean, x=species, fill = species)) +
409 | geom_bar(stat="identity")+
410 | geom_errorbar(aes(ymin = mean-sd, ymax = mean+sd),
411 | width=.1)
412 |
413 | ```
414 |
415 |
416 | Create the plot like the code above for the scatter plot.
417 |
418 | ```{r, warning=FALSE, fig.align = "center", out.width = '50%'}
419 |
420 | bar <- ggplot(penguin_summary, aes(y = mean, x = species, fill = species)) +
421 | geom_bar(stat = "identity")+
422 | geom_errorbar(aes(ymin = mean-sd, ymax = mean+sd),
423 | width=.1)+
424 | theme_bw(base_size = 20)+
425 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
426 | labs(x = "Species", y = "Body Mass (g)")+
427 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
428 | theme(legend.position = "none") +
429 | scale_y_continuous(limits = c(0, 5700))+
430 | theme(axis.title.y = element_text(vjust = 3)) +
431 | theme(axis.title.x = element_text(vjust = -1))
432 |
433 | bar
434 |
435 | ```
436 |
437 |
438 |
439 | ### Histogram
440 |
441 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
442 |
443 | ggplot(penguins, aes(x = flipper_length_mm, fill = species)) +
444 | geom_histogram(alpha = 0.4)+
445 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))
446 |
447 | ```
448 |
449 | ### Density Plot
450 |
451 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
452 |
453 | ggplot(penguins, aes(x = flipper_length_mm, fill = species)) +
454 | geom_density(alpha = 0.4)+
455 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))
456 |
457 | ```
458 |
459 |
460 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
461 |
462 | flipper <- ggplot(penguins, aes(x = flipper_length_mm, fill = species, colour = species)) +
463 | geom_density(alpha = 0.4)+
464 | scale_fill_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
465 | scale_colour_manual(values = c( "#FF6A00","#C15CCB", "#00868B"))+
466 | theme_bw(base_size = 20)+
467 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
468 | labs(x = "Flipper Length (mm)", y = "Density", fill = "Species", colour = "Species")+
469 | #theme(legend.position = "none") +
470 | scale_x_continuous(expand = c(0,0), limits = c(165, 238))+
471 | scale_y_continuous(expand = c(0,0), limits = c(0, 0.065), breaks=seq(0, 0.065, 0.02))+
472 | theme(axis.title.y = element_text(vjust = 3)) +
473 | theme(axis.title.x = element_text(vjust = -1))
474 |
475 | flipper
476 |
477 | ```
478 |
479 | ### Add a mean line
480 |
481 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
482 |
483 | flipper_mean <- penguins %>%
484 | group_by(species) %>%
485 | summarise(mean = mean(flipper_length_mm, na.rm = TRUE))
486 |
487 | density <- flipper +
488 | geom_vline(data = flipper_mean, aes(xintercept = mean, colour = species), linetype = "dashed", size = 1.3)
489 |
490 | density
491 | ```
492 |
493 |
494 |
495 |
496 | ### Box plot
497 |
498 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
499 |
500 |
501 | ggplot(na.omit(penguins), aes(x=species, y=flipper_length_mm, fill=sex)) +
502 | geom_boxplot()
503 | ```
504 |
505 |
506 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, out.width = '50%'}
507 |
508 | box_plot<- ggplot(na.omit(penguins), aes(x=species, y=flipper_length_mm, fill=sex)) +
509 | geom_boxplot()+
510 | theme_bw(base_size = 20)+
511 | theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())+
512 | labs(x = "Species", y = "Flipper Length (mm)", fill = "Sex")+
513 | theme(axis.title.y = element_text(vjust = 3)) +
514 | theme(axis.title.x = element_text(vjust = -1)) +
515 | theme(legend.position = c(0.15, 0.8))
516 |
517 | box_plot
518 | ```
519 |
520 |
521 |
522 | ### Arrange plots
523 |
524 | There are several different methods that allow you to arrange your plots into different panels.
525 |
526 | #### `ggarrange`
527 |
528 | Here we will arrange four of the plots that we made above into one figure using `ggarrange`. You can specify the number of rows `nrow` and `ncol` and add `labels.
529 |
530 | Note we previously saved our figures (`ellipse`, `density`, `bar`, `box_plot`) in the Global Environment and we are calling on them here.
531 |
532 |
533 | ``````{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=12, fig.height=8}
534 | library(ggpubr)
535 |
536 | ggarrange(ellipse, density, bar, box_plot,
537 | nrow = 2, ncol = 2,
538 | labels = c("a", "b", "c", "d"))
539 |
540 | ```
541 |
542 | #### Same axes
543 |
544 | If two plots have the same axes, you can remove one and use `align = "v"` to ensure the plots remain the same size.
545 |
546 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=12, fig.height=8}
547 |
548 | # Remove y axes text and title
549 | lm <- lm +
550 | theme(axis.title.y=element_blank(),
551 | axis.text.y=element_blank(),
552 | axis.ticks.y=element_blank())
553 |
554 |
555 | ggarrange(ellipse, lm,
556 | nrow = 1, ncol = 2,
557 | labels = c("a", "b"),
558 | align = "v")
559 |
560 | ```
561 |
562 | #### `patchwork`
563 |
564 | `patchwork` is a very straighforward and intuative package for arranging plots.
565 | e.g:
566 |
567 | - `plot1 + plot2` aligns two plots next to each other
568 | - `plot1 / plot2` aligns plot2 under plot1
569 |
570 | Lets make a fancy one.
571 |
572 | We add the figure labels with e.g. `labs(tag = 'a')`. We will also make two of our plots half the size of the main plot with `plot_layout(widths=c(2,1))`.
573 |
574 |
575 | ```{r, warning=FALSE, fig.align = "center", message = FALSE, fig.width=18, fig.height=10}
576 |
577 | library(patchwork)
578 |
579 | (ellipse + labs(tag = 'a')| ((bar+ labs(tag = 'b')) / (box_plot +labs(tag = 'c')))) + plot_layout(widths=c(2,1))
580 |
581 | ```
582 |
583 |
--------------------------------------------------------------------------------
/ggplot2-for-publications.Rproj:
--------------------------------------------------------------------------------
1 | Version: 1.0
2 |
3 | RestoreWorkspace: Default
4 | SaveWorkspace: Default
5 | AlwaysSaveHistory: Default
6 |
7 | EnableCodeIndexing: Yes
8 | UseSpacesForTab: Yes
9 | NumSpacesForTab: 2
10 | Encoding: UTF-8
11 |
12 | RnwWeave: Sweave
13 | LaTeX: pdfLaTeX
14 |
--------------------------------------------------------------------------------