How to specify different line types using geom_smooth

Question

I am making a graph for my data and am using the geom_smooth function in ggplot to make them smooth looking line graphs. I would like to make the graphs colorblind friendly, so would like to make the lines either different line types or the points different shapes. The graphs do not like the code that I am using and displays this graph:

Graph output

Here is the code that I am using:

ggplot(leach.conc, aes(x = days, y = cl_conc, linetype = treatment, color = factor(treatment, level = c('Liquid', 'Solid', 'KCl', 'Control')))) +
  labs(x = "Days since application", y = "Chloride concentration (mg/L)", color = "Treatment") +
  geom_smooth() + facet_wrap(.~soil_type) +
  scale_x_continuous(breaks = c(0,4,11,18))

Here is a dput for reproducibility

leach.conc <- data.frame(
  core = c(
    "MS", "MS", "MS", "MS", "ML", "ML", "ML", "ML", "MK", "MK",
    "MK", "MK", "MC", "MC", "MC", "MC", "FS", "FS", "FS", "FS", "FL",
    "FL", "FL", "FL", "FK", "FK", "FK", "FK", "FC", "FC", "FC", "FC",
    "MS", "MS", "MS", "MS", "ML", "ML", "ML", "ML", "MK", "MK", "MK",
    "MK", "MC", "MC", "MC", "MC", "FS", "FS", "FS", "FS", "FL", "FL",
    "FL", "FL", "FK", "FK", "FK", "FK", "FC", "FC", "FC", "FC", "MS",
    "MS", "MS", "MS", "ML", "ML", "ML", "ML", "MK", "MK", "MK", "MK",
    "MC", "MC", "MC", "MC", "FS", "FS", "FS", "FS", "FL", "FL", "FL",
    "FL", "FK", "FK", "FK", "FK", "FC", "FC", "FC", "FC", "CS", "CL",
    "CK", "CC", "PS", "PL", "PK", "PC", "CS", "CL", "CK", "CC", "PS",
    "PL", "PK", "PC", "CS", "CL", "CK", "CC", "PS", "PL", "PK", "PC",
    "CS", "CL", "CK", "CC", "PS", "PL", "PK", "PC", "CS", "CL", "CK",
    "CC", "PS", "PL", "PK", "PC", "CS", "CL", "CK", "CC", "PS", "PL",
    "PK", "PC", "CS", "CL", "CK", "CC", "PS", "PL", "PK", "PC", "CS",
    "CL", "CK", "CC", "PS", "PL", "PK", "PC", "CS", "CL", "CK", "CC",
    "PS", "PL", "PK", "PC", "CS", "CK", "CL", "CC", "PS", "PL", "PK",
    "PC", "CS", "CL", "CK", "CC", "PS", "PL", "PK", "PC", "CS", "CL",
    "CK", "CC", "PS", "PL", "PK", "PC"
  ),
  core_id = c(
    "MS1", "MS1", "MS1", "MS1", "ML1", "ML1", "ML1", "ML1", "MK1",
    "MK1", "MK1", "MK1", "MC1", "MC1", "MC1", "MC1", "FS1", "FS1",
    "FS1", "FS1", "FL1", "FL1", "FL1", "FL1", "FK1", "FK1", "FK1",
    "FK1", "FC1", "FC1", "FC1", "FC1", "MS2", "MS2", "MS2", "MS2",
    "ML2", "ML2", "ML2", "ML2", "MK2", "MK2", "MK2", "MK2", "MC2",
    "MC2", "MC2", "MC2", "FS2", "FS2", "FS2", "FS2", "FL2", "FL2",
    "FL2", "FL2", "FK2", "FK2", "FK2", "FK2", "FC2", "FC2", "FC2",
    "FC2", "MS3", "MS3", "MS3", "MS3", "ML3", "ML3", "ML3", "ML3",
    "MK3", "MK3", "MK3", "MK3", "MC3", "MC3", "MC3", "MC3", "FS3",
    "FS3", "FS3", "FS3", "FL3", "FL3", "FL3", "FL3", "FK3", "FK3",
    "FK3", "FK3", "FC3", "FC3", "FC3", "FC3", "CS1", "CL1", "CK1",
    "CC1", "PS1", "PL1", "PK1", "PC1", "CS2", "CL2", "CK2", "CC2",
    "PS2", "PL2", "PK2", "PC2", "CS3", "CL3", "CK3", "CC3", "PS3",
    "PL3", "PK3", "PC3", "CS1", "CL1", "CK1", "CC1", "PS1", "PL1",
    "PK1", "PC1", "CS2", "CL2", "CK2", "CC2", "PS2", "PL2", "PK2",
    "PC2", "CS3", "CL3", "CK3", "CC3", "PS3", "PL3", "PK3", "PC3",
    "CS1", "CL1", "CK1", "CC1", "PS1", "PL1", "PK1", "PC1", "CS2",
    "CL2", "CK2", "CC2", "PS2", "PL2", "PK2", "PC2", "CS3", "CL3",
    "CK3", "CC3", "PS3", "PL3", "PK3", "PC3", "CS1", "CK1", "CL1",
    "CC1", "PS1", "PL1", "PK1", "PC1", "CS2", "CL2", "CK2", "CC2",
    "PS2", "PL2", "PK2", "PC2", "CS3", "CL3", "CK3", "CC3", "PS3",
    "PL3", "PK3", "PC3"
  ),
  soil_type = rep(
    c(
      "WSL", "NCL ", "WSL", "NCL ", "WSL", "NCL ", "ESL", "FAH",
      "ESL", "FAH", "ESL", "FAH", "ESL", "FAH", "ESL", "FAH", "ESL",
      "FAH", "ESL", "FAH", "ESL", "FAH", "ESL", "FAH", "ESL", "FAH",
      "ESL", "FAH", "ESL", "FAH"
    ),
    rep(c(16L, 4L), c(6L, 24L))
  ),
  treatment = c(
    "Tl", "Tl", "Tl", "Tl", "SM", "SM", "SM", "SM", "KCl", "KCl",
    "KCl", "KCl", "Control", "Control", "Control", "Control", "Tl",
    "Tl", "Tl", "Tl", "SM", "SM", "SM", "SM", "KCl", "KCl", "KCl",
    "KCl", "Control", "Control", "Control", "Control", "Tl", "Tl",
    "Tl", "Tl", "SM", "SM", "SM", "SM", "KCl", "KCl", "KCl", "KCl",
    "Control", "Control", "Control", "Control", "Tl", "Tl", "Tl",
    "Tl", "SM", "SM", "SM", "SM", "KCl", "KCl", "KCl", "KCl", "Control",
    "Control", "Control", "Control", "Tl", "Tl", "Tl", "Tl", "SM",
    "SM", "SM", "SM", "KCl", "KCl", "KCl", "KCl", "Control", "Control",
    "Control", "Control", "Tl", "Tl", "Tl", "Tl", "SM", "SM", "SM",
    "SM", "KCl", "KCl", "KCl", "KCl", "Control", "Control", "Control",
    "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control",
    "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl",
    "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM",
    "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl",
    "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control",
    "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl",
    "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM",
    "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl",
    "Control", "Tl", "KCl", "SM", "Control", "Tl", "SM", "KCl", "Control",
    "Tl", "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control", "Tl",
    "SM", "KCl", "Control", "Tl", "SM", "KCl", "Control"
  ),
  days = c(
    0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L,
    11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L,
    0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L,
    11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L,
    0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L,
    11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L,
    0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 4L, 11L, 18L, 0L, 0L,
    0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
    0L, 0L, 0L, 0L, 0L, 0L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L,
    4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 11L,
    11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L,
    11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 11L, 18L, 18L, 18L,
    18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L,
    18L, 18L, 18L, 18L, 18L, 18L, 18L, 18L
  ),
  cl_conc = c(
    3.6, 18.1, 18.1, 17.4, 3.8, 77.1, 81.4, 66.8, 4.6, 19.4, 22.3, 36.9, 4.1, 1.9,
    1.2, 0.6, 18.2, 27.8, 28.3, 28.3, 15, 107.8, 150.3, 94.6, 12.3, 84.8, 53.4,
    51.9, 17.3, 9.1, 4.25, 1.9, 3.4, 19.8, 20.7, 20.5, 2.3, 102, 56.7, 47.4, 2.6,
    33.4, 15.3, 19.9, 2.9, 2, 1.2, 0.8, 8.1, 37.1, 39.8, 34.8, 4.9, 81.9, 67.5,
    56, 5.9, 41.1, 38.3, 30.9, 17.5, 12.4, 6, 3.1, 2.4, 27.8, 27.8, 24.9, 2.6,
    79.7, 65.5, 55.2, 2.1, 13.5, 20.4, 24.7, 2, 1.6, 1.2, 0.7, 8.4, 42.7, 40.5,
    30.1, 10.3, 121.2, 73.6, 38, 9.4, 53, 38.5, 22.3, 6.6, 4.7, 1.9, 0.85, 4.3,
    5.5, 5.6, 7.3, 3.7, 0.9, 2.8, 7.633503186, 5.245095388, 4.9, 6.3, 5.6, 1.4, 5,
    5.6, 1.4, 5.7, 15.80504568, 7.6, 6.5, 2.5, 2, 2.6, 1.6, 46.5, 52.6, 32.9, 2.8,
    45.1, 1.3, 51.2, 2.6, 47.59251129, 68.3, 38.8, 5.4, 34.1, 66.7, 23.51266468,
    0.6, 34.2, 55.7, 23.8, 5, 42.1, 47.9, 44.3, 0.8, 56.23151874, 81.2, 36.1, 1.6,
    36.3, 48.2, 35.6, 1.5, 44.8, 80.9, 34.66600908, 3.1, 33.3, 81.5, 20.2, 0.4,
    40.1, 66.8, 24.5, 3.6, 39, 68.2, 36, 0.303367677, 31.1, 23.2, 75.7, 0.6, 26.2,
    45.3, 21.3, 0.6, 33.76030379, 47.5, 20.5, 1.1, 28.6, 65.9, 18.9, 0.2, 30.2,
    65.5, 23.3, 2.7, 23.9, 64, 24.7, 0.1
  ),
  cl_load = c(
    0.0072, 0.058825, 0.0543, 0.0609, 0.0076, 0.26985, 0.26455, 0.2171, 0.0092,
    0.0582, 0.0669, 0.119925, 0.0082, 0.0057, 0.0036, 0.00195, 0.0364, 0.09035,
    0.0849, 0.0849, 0.03, 0.3773, 0.4509, 0.3311, 0.0246, 0.2756, 0.1602, 0.1557,
    0.0346, 0.03185, 0.010625, 0.006175, 0.0034, 0.06435, 0.07245, 0.07175,
    0.00345, 0.3315, 0.19845, 0.1659, 0.0039, 0.10855, 0.05355, 0.064675,
    0.003625, 0.0065, 0.0039, 0.0026, 0.01215, 0.1113, 0.1393, 0.1044, 0.00735,
    0.266175, 0.185625, 0.168, 0.00885, 0.113025, 0.105325, 0.0927, 0.02625,
    0.0372, 0.018, 0.010075, 0.006, 0.09035, 0.09035, 0.0747, 0.0065, 0.2391,
    0.22925, 0.1656, 0.004725, 0.0405, 0.0663, 0.06175, 0.0055, 0.0048, 0.0036,
    0.0021, 0.0252, 0.1281, 0.131625, 0.097825, 0.0309, 0.3939, 0.2392, 0.114,
    0.0235, 0.17225, 0.125125, 0.0669, 0.0165, 0.015275, 0.006175, 0.0023375,
    0.0043, 0.00825, 0.0042, 0.0073, 0.0037, 0.001575, 0.0014, 0.011450255,
    0.005245095, 0.0049, 0.007875, 0.0042, 0.00175, 0.005, 0.0112, 0.0014,
    0.00285, 0.015805046, 0.0114, 0.008125, 0.00375, 0.0025, 0.00325, 0.0032,
    0.11625, 0.14465, 0.0987, 0.0084, 0.0902, 0.0039, 0.1152, 0.00715,
    0.118981278, 0.2049, 0.1164, 0.0135, 0.093775, 0.2001, 0.064659828, 0.0018,
    0.0855, 0.1671, 0.06545, 0.0125, 0.094725, 0.131725, 0.099675, 0.0024,
    0.177129284, 0.2436, 0.1083, 0.0048, 0.1089, 0.15665, 0.1068, 0.0045, 0.1344,
    0.2427, 0.11266453, 0.010075, 0.0999, 0.2445, 0.0606, 0.0012, 0.1203, 0.2004,
    0.0735, 0.0126, 0.117, 0.2046, 0.117, 0.000985945, 0.0933, 0.0696, 0.2268,
    0.0018, 0.0917, 0.15855, 0.0639, 0.00195, 0.101280911, 0.1425, 0.0615,
    0.003025, 0.0858, 0.214175, 0.0567, 0.00065, 0.1057, 0.212875, 0.075725,
    0.0081, 0.0717, 0.192, 0.0741, 3e-04
  )
)

You're getting two different legends because you renamed the `color` legend to `Treatment` so it can't be merged with `treatment` (lowercase T). Also you have a typo because the values of `treatment` you list in your `factor()` call don't match what you put in the data you shared. But basically you should do exactly what @zephryl said and use a colorblind-friendly palette. The different `linetype`s are pretty subtle. — Dan Adams, Apr 07 '23 at 21:28
@zephryl, yes the issue is that there are 2 legends and the line types are too similar (I really thought they were pretty much the same). When I try to do the transformation earlier, then the treatments are out of order in the legend. The only way I have been able to successfully get them to be in that order, is by doing the transformation in the plotting code. — mbelanger081, Apr 07 '23 at 21:31

zephryl · Answer 1 · 2023-04-07T22:01:05.180

Seems there are a few issues:

Separate legends. You're getting different legends for linetype and color because (1) in aes(), you transform treatment when passing it to color but not to linetype, and (2) you set the label for color to "Treatment", but the label for linetype is implicitly set to "treatment". The easiest way to avoid both these issues is to transform and rename before plotting, and just use the new variable in the plot spec:

leach.conc$Treatment <- factor(
  leach.conc$treatment, 
  levels = c("Tl", "SM", "KCl", "Control")
)

Similar-looking lines. You can specify more distinct-looking linetypes using scale_linetype_manual(). Specify lines as a quoted two digit number, where the first digit is the length of each solid segment and the second is the length of each gap. e.g., "51" gives you long dashes with short gaps, while "15" will give you short dashes with long gaps.

Non-colorblind-safe colors. You can view colorblind-safe Brewer palettes using RColorBrewer::display.brewer.all(colorblindFriendly = TRUE). If we specify only qualitative palettes using RColorBrewer::display.brewer.all(colorblindFriendly = TRUE, type = "qual"), we get:

Putting it all together:

library(ggplot2)

ggplot(
    leach.conc, 
    aes(x = days, y = cl_conc, linetype = Treatment, color = Treatment, fill = Treatment)
  ) +
  geom_smooth(alpha = .15) + 
  facet_wrap(.~soil_type) +
  scale_x_continuous(breaks = c(0,4,11,18)) +
  scale_linetype_manual(values = c("51", "34", "11", "solid")) +
  scale_color_brewer(palette = "Dark2", aesthetics = c("color", "fill")) +
  labs(x = "Days since application", y = "Chloride concentration (mg/L)") +
  theme_light()

other useful resources for colour-blind palettes are colorbrewer2.org or the [colorblindr package (not on CRAN)](https://github.com/clauswilke/colorblindr). It features scale_ functions based on the "Okabe-Ito" colors, which is also the default palette when calling `palette.colors()` — tjebo, Apr 08 '23 at 11:03

How to specify different line types using geom_smooth

1 Answers1