GG plot legend stacking all shapes / deleting one item from legend

user_15633943 0
Viewed 65

There is a problem with my ggplot, I want to have different shapes for certain values. But in the legend all thing stack on top of each other and therefore the legend is not clear anymore. Moreover, I would like to delete Japan from the legend or find another way to make it red and shape=17. Maybe add to the filter Japan=='FALSE'? I tried but did not succeed...

Here is sample of my data:

    structure(list(Country = c("Albania", "Aruba", "Austria", "Barbados", 
"Bosnia and Herzegovina", "Canada", "China, Hong Kong SAR", "China, Macao SAR", 
"Croatia", "Curaçao", "Denmark", "Finland", "France", "Germany", 
"Iceland", "Italy", "Japan", "Latvia", "Lithuania", "Malta", 
"Mauritius", "Montenegro", "Netherlands", "New Zealand", "Poland", 
"Portugal", "Republic of Korea", "Serbia", "Singapore", "Slovenia", 
"Sri Lanka", "Taiwan", "Thailand", "Trinidad and Tobago", "United States of America"
), `Dependency Ratio 1990` = c(0.371731839842905, 0.42945960478559, 
0.698167620530499, 0.444513116903726, 0.511357742868368, 0.519783119456753, 
0.444426949479237, 0.30306654331295, 0.723691486939267, 0.424414908111054, 
0.68769508504734, 0.641530173960242, 0.690189226564259, 0.755969184286434, 
0.520917100019657, 0.763735128335739, 0.692461922514607, 0.728970209495916, 
0.655093765838824, 0.556158238426314, 0.308439455191019, 0.551893405455789, 
0.582543266573117, 0.548269437314668, 0.592240027149362, 0.744368260326749, 
0.33818760118961, 0.653157768845158, 0.294237762460344, 0.611402526341597, 
0.354595845574429, 0.391092962761626, 0.331304119150256, 0.35111793456609, 
0.562804979721953), `Average Age 1990` = c(40.3688042387203, 
42.5004114258846, 46.6904752788518, 42.5683625031078, 42.4530074518545, 
44.2409448871874, 42.0677766503007, 39.7173235436725, 46.2329924328207, 
42.2501753565583, 47.1375106133558, 46.3380103826365, 46.6915593676301, 
46.9202073747455, 44.129974503284, 47.1071528898825, 46.6077408054755, 
46.4664135824761, 45.460688263743, 44.9450928096016, 39.2332051727974, 
43.66848, 45.1863467813393, 44.5466909246095, 44.9318462263063, 
46.8407998745322, 39.6873706785703, 45.3128111624097, 39.2982502106955, 
45.1205082490539, 40.2124158913374, 40.9051762916043, 39.4534335710941, 
40.4173693037492, 45.0904477728946), ...4 = c(NA, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), 
    `Dependency Ratio 2000` = c(0.457221087782508, 0.45940989018547, 
    0.742565467519652, 0.541204030550029, 0.597208883500012, 
    0.622668782217446, 0.473107405821069, 0.344264744501091, 
    0.780230513979289, 0.633789601501269, 0.790840886898238, 
    0.816743961984905, 0.766496601277572, 0.821136017787255, 
    0.572251283384235, 0.849390138872188, 0.927739183233871, 
    0.791488299481733, 0.704545225683664, 0.702609326498817, 
    0.35385418751795, 0.612564625368555, 0.69550083971213, 0.617845149047375, 
    0.611138887992547, 0.758176723785889, 0.399735388267277, 
    0.715467873467691, 0.383896159972764, 0.671137540638121, 
    0.407831309113246, 0.419443507121452, 0.374126385687095, 
    0.409593048372564, 0.615930392620661), `Average Age 2000` = c(42.9309383891972, 
    43.8674007980144, 47.739334648896, 45.3560289004102, 45.2990249348384, 
    46.3058678455289, 44.3575197674921, 42.5567755821042, 47.8706196243093, 
    46.6926342578517, 47.9056748231027, 48.2912968951969, 48.1201704908476, 
    48.7598382100637, 45.3475147626354, 48.7908038019529, 48.7739160208226, 
    47.837540150878, 46.9150297452015, 46.5384376276976, 41.1677637838199, 
    45.6830176554619, 46.9484767952653, 46.0162750047118, 45.6870193241911, 
    47.5083105450284, 41.9472090972845, 46.7580169116961, 42.3830413567395, 
    46.6274879755993, 41.6098644987726, 42.77485916275, 41.5907978667698, 
    41.9719792296039, 46.2781534087236), ...7 = c(NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA), `Dependency Ratio 2015` = c(0.769855001031037, 0.896573728647162, 
    0.952116351996821, 0.847635802309437, 0.907658891504387, 
    0.913339561634508, 0.877436210055064, 0.597279210234922, 
    1.04631340359464, 1.01174318826707, 0.977891738355926, 1.08472890326446, 
    1.01372962185931, 1.12997699958302, 0.77882885539859, 1.08347132170333, 
    1.20679093156161, 0.990147667477283, 0.994357961860711, 0.950194923573131, 
    0.666562280517562, 0.828817278052088, 0.989460150334804, 
    0.831814166715077, 0.847326002560978, 1.03326726133893, 0.749863481391909, 
    0.917215066264046, 0.671774028057953, 0.971441198307662, 
    0.611354032233621, 0.748057645284422, 0.665957813686028, 
    0.602213503073687, 0.855245238291093), `Average Age 2015` = c(46.3902100558352, 
    47.9818955923079, 49.878498965043, 48.4578763127188, 48.3870694416244, 
    48.7063314226308, 48.5131007402609, 44.1281261495054, 50.522561636728, 
    49.6455293947711, 49.9810732770387, 50.7119476819108, 50.6805196046482, 
    51.5236122201751, 47.3079677856577, 51.8838669025279, 53.2344169277342, 
    50.0674941000466, 49.9057070057583, 49.1748722211516, 45.3413745873924, 
    47.6617051653597, 49.9107746561504, 48.0022465682781, 48.1684244717051, 
    51.1526322354916, 47.160655712273, 49.0269050604693, 45.5481140676913, 
    50.0264456515826, 44.8882173741791, 47.0138589294768, 46.131374630996, 
    44.5204789350954, 48.0998439723386), `rgdpe 1990` = c(12005.7568359375, 
    2575.25561523438, 208007.234375, 4099.8515625, 6946.330078125, 
    915724.6875, 151044.28125, 9127.78125, 64448.71484375, NA, 
    144018.203125, 136194.359375, 1581529.625, 2204488.5, 8510.248046875, 
    1560881.5, 3552613.25, 44957.03515625, 53623.6875, 5354.54541015625, 
    11257.8095703125, 6702.1552734375, 427072.25, 76859.65625, 
    335254.875, 157535.140625, 565140.75, 113435.8046875, 64860.5703125, 
    42714.70703125, 55257.37109375, 430917.25, 308367.4375, 15085.6611328125, 
    9847675), `pop 1990` = c(3.286073, 0.062149, 7.723949, 0.260936, 
    4.463423, 27.541319, 5.727938, 0.343808, 4.776374, NA, 5.141115, 
    4.996222, 58.235697, 79.053984, 0.255043, 57.048236, 124.50524, 
    2.664439, 3.696035, 0.362015, 1.055868, 0.615002, 14.965448, 
    3.398172, 37.960193, 9.895364, 42.918419, 9.517675, 3.012966, 
    2.006405, 17.325773, 20.278946, 56.558186, 1.221116, 252.120309
    ), `emp 1990` = c(1.32407820224762, NA, 3.56034135818481, 
    0.105200000107288, 1.68987882137299, 13.2902002334595, 2.73075985908508, 
    0.16329999268055, 2.17813229560852, NA, 2.63417220115662, 
    2.47324681282043, 23.6595039367676, 39.5477294921875, 0.138074412941933, 
    22.8031978607178, 65.1040191650391, 1.25425291061401, 1.70560574531555, 
    0.132750615477562, 0.403737008571625, 0.174824863672256, 
    6.80782461166382, 1.52131986618042, 15.0829668045044, 4.46721506118774, 
    18.2060832977295, 4.61394643783569, 1.52955627441406, 1.1292530298233, 
    5.04270553588867, 8.64918994903564, 28.7045097351074, 0.374099999666214, 
    123.046020507812), ...5 = c(NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), `rgdpe 2000` = c(15180.880859375, 
    4031.13427734375, 314579.6875, 6460.9755859375, 21482.595703125, 
    1276875.5, 256729.140625, 13199.31640625, 58379.2734375, 
    NA, 203091.515625, 186055.46875, 2135621.25, 3030253, 11653.5791015625, 
    2081385.625, 4696670.5, 26473.306640625, 42346.78125, 9706.8212890625, 
    18593.318359375, 4850.50634765625, 691869.625, 114914.1171875, 
    563679.1875, 259576.953125, 1150272, 60002.0703125, 166273.53125, 
    48606.53125, 90442.875, 789527.25, 504829.28125, 18303.63671875, 
    14110581), `pop 2000` = c(3.129243, 0.090853, 8.069276, 0.271515, 
    3.751176, 30.588383, 6.606327, 0.427782, 4.428075, NA, 5.341194, 
    5.187954, 60.874357, 81.400882, 0.280435, 56.692178, 127.524174, 
    2.384164, 3.501839, 0.393645, 1.185145, 0.613559, 15.926188, 
    3.858999, 38.556693, 10.297112, 47.379241, 7.516346, 4.028871, 
    1.987717, 18.777601, 22.18453, 62.952642, 1.267153, 281.710909
    ), `emp 2000` = c(0.962967455387115, 0.0419000014662743, 
    3.7599310874939, 0.12899999320507, 0.643303751945496, 14.952766418457, 
    3.20262169837952, 0.195299997925758, 1.67029082775116, NA, 
    2.75595617294312, 2.30501818656921, 25.6252250671387, 39.6031150817871, 
    0.1570855230093, 22.91796875, 65.9155044555664, 0.930018603801727, 
    1.40124833583832, 0.146938025951385, 0.464872002601624, 0.176752656698227, 
    8.20334815979004, 1.81842231750488, 14.4786930084229, 5.076171875, 
    21.4411239624023, 3.0847954750061, 2.08465480804443, 0.917375922203064, 
    6.30462980270386, 9.59665679931641, 31.47385597229, 0.503100037574768, 
    138.636108398438), ...9 = c(NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), `rgdpe 2015` = c(32037.935546875, 
    3959.59252929688, 448794.71875, 4856.21044921875, 40599.22265625, 
    1659691.75, 411350.65625, 62493.578125, 99181.7265625, 4041.00463867188, 
    278112.53125, 237412.921875, 2772463.25, 3915258.25, 16865.345703125, 
    2296760.75, 5094436, 51517.390625, 87529.2890625, 17455.76171875, 
    25619.560546875, 10165.248046875, 872643.75, 174613.65625, 
    1069768.375, 314019.625, 1928056.875, 108470.875, 451476.4375, 
    68875.1796875, 242116.15625, 1125999, 1108115.875, 38140.46484375, 
    18905122), `pop 2015` = c(2.890513, 0.104341, 8.67866, 0.285324, 
    3.429361, 36.026676, 7.185996, 0.602085, 4.232874, 0.159847, 
    5.688695, 5.481122, 66.596315, 81.787411, 0.330243, 60.578494, 
    127.985133, 1.997674, 2.93188, 0.433559, 1.259456, 0.626956, 
    16.938499, 4.614532, 38.034079, 10.368351, 50.823093, 7.095383, 
    5.592152, 2.071199, 20.908027, 23.462914, 68.714511, 1.370328, 
    320.87831), `emp 2015` = c(0.926395297050476, 0.0467174984514713, 
    4.27823972702026, 0.128199994564056, 0.616872131824493, 18.3558368682861, 
    3.77715754508972, 0.396699994802475, 1.69973313808441, 0.0617999993264675, 
    2.83158588409424, 2.52453279495239, 27.3850765228271, 42.5355796813965, 
    0.181162342429161, 24.4446144104004, 66.9830322265625, 0.897909104824066, 
    1.35488307476044, 0.191138163208961, 0.563370883464813, 0.221699982881546, 
    8.80725860595703, 2.36527323722839, 15.8249969482422, 4.60829973220825, 
    26.079252243042, 2.56693267822266, 3.65548992156982, 0.950674414634705, 
    7.83100032806396, 11.1978015899658, 37.9529876708984, 0.623300015926361, 
    150.248474121094), GDP_per_capita_1990 = c(3653.52712369369, 
    41436.7989064084, 26930.1667288326, 15712.0963090566, 1556.27868524337, 
    33249.1224367286, 26369.74793547, 26549.0659030622, 13493.2303968973, 
    NA, 28013.0289100711, 27259.4691298745, 27157.3915394195, 
    27885.8621470614, 33367.895009371, 27360.7320654051, 28533.8452421762, 
    16872.9834521451, 14508.4360672991, 14790.949021881, 10662.1372845019, 
    10897.7780128154, 28537.2178634412, 22617.9417198423, 8831.74843183753, 
    15920.0955745539, 13167.790500391, 11918.4364550691, 21527.1497628915, 
    21289.1749329024, 3189.31634933402, 21249.4894951641, 5452.2158383934, 
    12353.995142814, 39059.4277750151), GDP_per_capita_2015 = c(11083.8233721402, 
    37948.5775418759, 51712.4439429589, 17019.9858729681, 11838.7135843237, 
    46068.41191788, 57243.3739526156, 103795.274961177, 23431.2966940429, 
    25280.4534252872, 48888.6346077615, 43314.6574506096, 41630.8807777127, 
    47871.1601471283, 51069.5024667442, 37913.7974278463, 39804.9045274657, 
    25788.6875561278, 29854.3218216639, 40261.5600616064, 20341.7670382094, 
    16213.6546214966, 51518.3635810942, 37839.9491541071, 28126.5749855544, 
    30286.3613509998, 37936.6300079375, 15287.5292285138, 80733.9352542635, 
    33253.7721809927, 11580.0575659291, 47990.5863355251, 16126.3735835943, 
    27833.0916713006, 58916.7962147395), change_log_GDP_per_cap_1990_2015 = c(1.10979365523243, 
    -0.0879373277889428, 0.652451349776046, 0.0799574118764497, 
    2.02907746023446, 0.326099133412088, 0.775094455358737, 1.36342588458029, 
    0.551884486717116, NA, 0.556875228192604, 0.46309013522626, 
    0.427192960710386, 0.540393411165555, 0.425593287986558, 
    0.32620624489414, 0.332899198982464, 0.424222198470275, 0.721599336011078, 
    1.00138172800831, 0.645977367508866, 0.39729484866726, 0.590729193215088, 
    0.514621927745218, 1.15836185302803, 0.643115306926695, 1.05814340225474, 
    0.248950929776939, 1.32184407992377, 0.44596948530371, 1.28947786000817, 
    0.814672002339478, 1.08443394063506, 0.812246156576203, 0.411041939497887
    ), change_dependency_ratio_1990_2015 = c(0.398123161188132, 
    0.467114123861573, 0.253948731466323, 0.403122685405711, 
    0.396301148636019, 0.393556442177755, 0.433009260575827, 
    0.294212666921972, 0.322621916655375, 0.587328280156014, 
    0.290196653308586, 0.44319872930422, 0.323540395295048, 0.374007815296586, 
    0.257911755378933, 0.319736193367588, 0.514329009047002, 
    0.261177457981367, 0.339264196021887, 0.394036685146817, 
    0.358122825326543, 0.2769238725963, 0.406916883761686, 0.283544729400408, 
    0.255085975411616, 0.288899001012185, 0.411675880202299, 
    0.264057297418888, 0.377536265597609, 0.360038671966065, 
    0.256758186659192, 0.356964682522796, 0.334653694535773, 
    0.251095568507597, 0.29244025856914)), class = "data.frame", row.names = c(NA, 
-35L))

And this is my code:

#Adding OECD
#Exclude japan for fig
OECD <- c("Dem. People's Republic of Korea",'Mexico','Chile',
          'New Zealand','Czechia','Hungary','Slovakia','Denmark','Estonia',
          'Finland','Iceland','Ireland','Latvia','Lithuania','Norway',
          'Sweden','United Kingdom','Greece','Italy','Portugal','Slovenia',
          'Spain','Austria','Belgium','France','Germany','Luxembourg',
          'Netherlands','Switzerland','Australia','Canada','United States of America',
          'Poland','Turkey','Israel')

#Figure 2
library(tidyverse)
df  %>%
  mutate(OECD = factor(Country %in% OECD, labels = c("NonOECD","OECD"))) %>% mutate(Japan = factor(Country == 'Japan' , labels=c('FALSE','TRUE')))-> newdata

ggplot() +  
  geom_point(data = filter(newdata, OECD == 'NonOECD'),aes(x = change_dependency_ratio_1990_2015, y = change_log_GDP_per_cap_1990_2015, colour='NonOECD'),shape = 16, size=3) +
   geom_point(data = filter(newdata, Japan == 'TRUE'),aes(x = change_dependency_ratio_1990_2015, y = change_log_GDP_per_cap_1990_2015,colour='Japan'),shape = 17, size=4) +
  geom_point(data = filter(newdata, OECD == 'OECD'),aes(x = change_dependency_ratio_1990_2015, y = change_log_GDP_per_cap_1990_2015, colour='OECD' ),shape = 18, size=4) + 
  
  scale_color_manual(values = c(NonOECD = "cyan4", OECD = "orange",Japan='red'),
                     labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries")) + geom_abline(size=1, col='grey')+
  
  theme_classic()+ theme( panel.grid.major.y =element_line(color='grey', size=0.7),legend.title = element_blank(),
                          panel.grid.minor.y =element_blank(), 
                          legend.background = element_blank(), legend.box.background = element_rect(colour = "black"),
                          legend.spacing.y = unit(0, "mm"),legend.direction = 'horizontal',
                          legend.position = "bottom",aspect.ratio = 0.7, axis.text = element_text(colour = 1, size = 13),)

Example of what I got: enter image description here

And what it should look like: enter image description here

Thanking you in advance!!

ggplot2 legend r
1 Answers

The code below produces a plot equivalent to the expected output.
The two main differences are:

  1. There is no data for "Japan" in the question so I have substituted "Portugal" (my country) for it;
  2. There is no GDP data, logged or not so I have created a new column with random uniform numbers, runif.

The plot is in fact simple, to create the factor OECD start by creating a logical vector, then use an ifelse to assign an integer value to the special country, in this case "Portugal" and add 2 to the other logical vector's elements, giving FALSE/TRUE + 2 == 0/1 + 2.

In order not to mix the plot with the theme, I have also created a custom theme, with code at the end.

library(tidyverse)

set.seed(2021)

df %>% 
  mutate(OECD = Location %in% OECD,
         OECD = ifelse(Location == "Portugal", 1L, OECD + 2L),
         OECD = factor(OECD, labels = c("Portugal", "NonOECD","OECD"))) %>% 
  mutate(GDP = runif(n(), -2, 2)) %>% 
  ggplot(aes(x = `Dependency Ratio`, y = GDP, color = OECD, shape = OECD, size = OECD)) +  
  geom_point() +
  scale_color_manual(
    values = c(NonOECD = "cyan4", OECD = "orange", Portugal = 'red'),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) + 
  scale_shape_manual(
    values = c(NonOECD = 16, OECD = 18, Portugal = 17),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  scale_size_manual(
    values = c(NonOECD = 4, OECD = 3, Portugal = 4),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  geom_abline(size = 1, col = 'grey') +
  theme_custom_Cas()

enter image description here

In order to remove the special country from the legend, subset the data. The code below will output the %>% pipe to a new data set and used it in the plot.

set.seed(2021)

df %>% 
  mutate(OECD = Location %in% OECD,
         OECD = ifelse(Location == "Portugal", 1L, OECD + 2L),
         OECD = factor(OECD, labels = c("Portugal", "NonOECD","OECD"))) %>% 
  mutate(GDP = runif(n(), -2, 2)) -> newdata

ggplot(newdata, aes(x = `Dependency Ratio`, y = GDP, color = OECD, shape = OECD, size = OECD)) +  
  geom_point(data = subset(newdata, OECD != "Portugal")) +
  # In the special country's layer the color, shape and size must be
  # outside aes() and show.legend = FALSE
  geom_point(
    data = subset(newdata, OECD == "Portugal"),
    color = "red", shape = 17, size = 4,
    show.legend = FALSE
  ) +
  scale_color_manual(
    values = c(NonOECD = "cyan4", OECD = "orange", Portugal = 'red'),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) + 
  scale_shape_manual(
    values = c(NonOECD = 16, OECD = 18, Portugal = 17),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  scale_size_manual(
    values = c(NonOECD = 4, OECD = 3, Portugal = 4),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  geom_abline(size = 1, col = 'grey') +
  theme_custom_Cas()

enter image description here

In order to have a fill color, the points shapes must be changed. See in help("points") the rightmost points, filled in grey. Those shapes allow for a border (ggplot aesthetic color) and a fill color (ggplot aesthetic fill).

ggplot(newdata, aes(x = `Dependency Ratio`, y = GDP, fill = OECD, shape = OECD, size = OECD)) +  
  geom_point(data = subset(newdata, OECD != "Portugal")) +
  # In the special country's layer the color, shape and size must be
  # outside aes() and show.legend = FALSE
  geom_point(
    data = subset(newdata, OECD == "Portugal"),
    fill = "red", shape = 24, size = 4,
    show.legend = FALSE
  ) +
  scale_fill_manual(
    values = c(NonOECD = "cyan4", OECD = "orange", Portugal = 'red'),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) + 
  scale_shape_manual(
    values = c(NonOECD = 21, OECD = 23, Portugal = 24),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  scale_size_manual(
    values = c(NonOECD = 4, OECD = 3, Portugal = 4),
    labels = c(NonOECD = "All Countries except OECD", OECD = "OECD countries", Portugal = "Portugal")
  ) +
  geom_abline(size = 1, col = 'grey') +
  theme_custom_Cas()

enter image description here

Custom theme code.

theme_custom_Cas <- function(){ 
  theme_classic() %+replace%    #replace elements we want to change
    theme(
      panel.grid.major.y = element_line(color = 'grey', size = 0.7),
      legend.title = element_blank(),
      panel.grid.minor.y = element_blank(), 
      legend.background = element_blank(), 
      legend.box.background = element_rect(colour = "black"),
      legend.spacing.y = unit(0, "mm"),
      legend.direction = 'horizontal',
      legend.position = "bottom",
      aspect.ratio = 0.7, 
      axis.text = element_text(colour = 1, size = 13)    
    )
}
user_82454060
Related