R ggplot double y-axis, log10 and continuous
Question
This question was closed for no reason, the provided link to a similar question does NOT provide an answer for this specific problem.
How can I have two different plots in one graph, with two separate y-axis, and with different scales?
df <- data.frame(x=3:22,
y1=c(3, 4, 5, 7, 9, 13, 15, 19, 23, 24, 29,
38, 40, 50, 56, 59, 70, 89, 104, 130),
y2=c(1,2,3,4,4,5,5,6,6,7,6,5,5,5,4,4,3,3,2,1))
ylim.prim <- c(min(df$y1), max(df$y1))
ylim.sec <- c(min(df$y2), max(df$y2))
b <- diff(ylim.prim)/diff(ylim.sec)
a <- ylim.prim[1] - b*ylim.sec[1]
ggplot(df, aes(x=x)) +
geom_line(aes(y=y1))+
geom_line(aes(y = a + y2*b))+
scale_y_continuous(sec.axis = sec_axis(~ (. - a)/b))
So with this, I can plot a double y-axis, but I would like either the left or the right one to be in log-scale. Does anyone have an idea how to do that? I would like to highlight, that each axis belongs to one of the two geom_lines.
exemplary real data:
df<-structure(list(x = c(0.3499125, 8.347913, 16.34591, 24.34391,
32.34191, 40.33992, 48.33792, 56.33592, 64.33392, 72.33192, 80.32992,
88.32792, 96.32592, 104.3239, 112.3219, 120.3199, 128.3179, 136.3159,
144.3139, 152.3119, 160.3099, 168.3079, 176.3059, 184.3039, 192.3019,
200.2999, 208.2979, 216.2959, 224.2939, 232.2919, 240.2899, 248.2879,
256.2859, 264.2839, 272.2819, 280.2799, 288.2779, 296.2759, 304.2739,
312.2719, 320.2699, 328.2679, 336.2659, 344.2639, 352.2619, 360.2599,
368.2579, 376.2559, 384.2539, 392.2519, 400.2451, 408.0882, 415.9314,
423.7745, 431.7164, 439.6766, 447.6368, 455.597, 463.5572, 471.5174,
479.4776, 487.4378, 495.398, 503.3582, 511.3184, 519.2786, 527.2465,
535.234, 543.2215, 551.209, 559.1966, 567.1841, 575.1716, 583.1591,
591.1466, 599.1342, 607.1217, 615.1092, 623.0967, 631.0842, 639.0718,
647.0593, 655.0468, 663.0343, 671.0218, 679.0094, 686.9969, 694.9844,
702.9719, 710.9594, 718.947, 726.9345, 734.922, 742.9095, 750.897,
758.8846, 766.8721, 774.8596, 782.8471, 790.8346, 798.8222, 806.8097,
814.7972, 822.7847, 830.7722, 838.7598, 846.7413, 854.7015, 862.6617,
870.6219, 878.5821, 886.5423, 894.5025, 902.4627, 910.4229, 918.3831,
926.3433, 934.3035, 942.2637, 949.9737, 957.7313, 965.6915, 973.6517,
981.6119, 989.5721, 997.5323, 1005.493, 1013.453, 1021.413, 1029.373,
1037.333, 1045.294, 1053.254, 1060.905, 1068.657, 1076.5, 1084.343,
1092.245, 1100.232, 1108.22, 1116.207, 1124.195, 1132.183, 1140.17,
1148.158, 1156.145, 1164.133, 1172.12, 1180.108, 1188.095, 1196.083,
1204.07, 1212.058, 1220.045, 1228.033, 1236.02, 1244.008, 1251.995,
1259.983, 1267.97, 1275.958, 1283.945, 1291.933, 1299.92, 1307.908,
1315.895, 1323.883, 1331.871, 1339.858, 1347.846, 1355.833, 1363.821,
1371.808, 1379.796, 1387.783, 1395.771, 1403.758, 1411.683, 1419.488,
1427.293, 1435.234, 1443.214, 1451.195, 1459.175, 1467.155, 1475.135,
1483.115, 1491.095, 1499.075, 1507.055, 1515.035, 1523.015, 1530.995,
1538.975, 1546.955, 1554.935, 1562.915, 1570.895, 1578.875, 1586.855,
1594.835, 1602.815, 1610.796, 1618.776, 1626.756), y2 = c(2.27,
2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27,
2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27,
2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27,
2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27, 2.27,
2.27, 2.27, 2.27, 2.27, 2.27, 2.300344, 2.300344, 2.300344, 2.300344,
2.356774, 2.356774, 2.356774, 2.356774, 2.356774, 2.356774, 2.356774,
2.356774, 2.356774, 2.356774, 2.356774, 2.356774, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672, 2.35672,
2.35672, 2.35672, 2.35672, 2.436441, 2.436441, 2.436441, 2.436441,
2.436441, 2.436441, 2.436441, 2.436441, 2.436441, 2.436441, 2.436441,
2.436441, 2.436441, 2.593907, 2.436372, 2.436372, 2.436372, 2.436372,
2.436372, 2.436372, 2.436372, 2.436372, 2.436372, 2.436372, 2.436372,
2.436372, 2.436372, 2.32186, 2.362217, 2.362217, 2.362217, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666, 2.356666,
2.356666, 2.356666, 2.356666, 2.356666, 2.6, 2.6, 2.6, 1.38,
1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38,
1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38, 1.38,
1.38, 1.38), y1 = c(1.0121711505181e-12, 1.36703949435522e-12,
1.84632494858934e-12, 2.49364868526704e-12, 3.36792573591631e-12,
4.54872499912722e-12, 6.14351261535188e-12, 8.29743536886833e-12,
1.12065270336094e-11, 1.5135549574501e-11, 2.04420900737953e-11,
2.76091094527446e-11, 3.7288908764166e-11, 5.03624430409839e-11,
6.80195634790675e-11, 9.1867308143082e-11, 1.24076096081818e-10,
1.67577290537946e-10, 2.26330068874485e-10, 3.05681854075198e-10,
4.12854243103824e-10, 5.57601379152127e-10, 7.53097024416113e-10,
1.01713383727135e-09, 1.37374208059794e-09, 1.85537854367786e-09,
2.50587585968424e-09, 3.38443865549292e-09, 4.57102650080046e-09,
6.17363418887535e-09, 8.33811836170825e-09, 1.12614720154805e-08,
1.52097587979964e-08, 2.05423179929812e-08, 2.77444811472099e-08,
3.74717262176886e-08, 5.06093638045714e-08, 6.83530636095683e-08,
9.23177269923811e-08, 1.24684445803943e-07, 1.68399029805712e-07,
2.2743990730289e-07, 3.07180534936803e-07, 4.14878418164378e-07,
5.60335245682253e-07, 7.56789448222779e-07, 1.02212079550612e-06,
1.38047758295943e-06, 1.86447536195262e-06, 2.51816121915984e-06,
3.40041627253216e-06, 4.54641422109325e-06, 6.03834395349185e-06,
7.97982308417383e-06, 1.04817057864209e-05, 1.35489394526282e-05,
1.72865813618296e-05, 2.18309175451225e-05, 2.73477089217644e-05,
3.40380929265256e-05, 4.21460976943755e-05, 5.1967369819244e-05,
6.38600965206997e-05, 7.82579725998248e-05, 9.56860600237062e-05,
0.000116779922497502, 0.000142335963852117, 0.000173355233081032,
0.000210921169220437, 0.00025641423771755, 0.000311506843394084,
0.000378223660039666, 0.000459016772258694, 0.000556855446703318,
0.000675335852469291, 0.000818812559327979, 0.000992558463515475,
0.00120295979079978, 0.00145774933544351, 0.00176629223196222,
0.00213992675332361, 0.00259238540452746, 0.00314030080012767,
0.00380380663276417, 0.00460729241649197, 0.00558028623723877,
0.00675855195676351, 0.00818539089553454, 0.00991324940277193,
0.0120056289887936, 0.0145394320140841, 0.0176077795085524, 0.0213234607142798,
0.0258230276512022, 0.0312718588796218, 0.0378702107462816, 0.0458605985746323,
0.0555366906666179, 0.0672541473478419, 0.0814435879234418, 0.0986265500083953,
0.119434540978074, 0.144632361864515, 0.175146169178876, 0.212097172372404,
0.256843809691636, 0.310804466617569, 0.369842684226363, 0.431179045399307,
0.493945429064933, 0.557253598745863, 0.620207670785615, 0.681916748818428,
0.741507890725135, 0.7981374131588, 0.85100435722966, 0.899360296843544,
0.942521141956581, 0.979876115698495, 0.999812477868897, 0.972735811573507,
0.932054481163834, 0.885815978946983, 0.834672310063189, 0.779345135479752,
0.720614966194481, 0.659310357439126, 0.596295772938628, 0.532460446628508,
0.468704750726898, 0.405928059993783, 0.345016117261179, 0.286827900234236,
0.237623456225718, 0.19790417419609, 0.164452220923679, 0.136482929335078,
0.112955983834528, 0.0932654285994524, 0.0770066613579728, 0.0635815895825471,
0.0524963384973958, 0.0433430973402779, 0.0357851408660264, 0.0295444442412585,
0.0243914258497928, 0.0201365021470287, 0.0166231553258446, 0.0137221430720713,
0.0113267390267136, 0.00934882441245559, 0.00771563465768068,
0.00636709119593569, 0.00525358045912998, 0.00433414349100376,
0.00357495316102798, 0.00294808230359634, 0.00243046968641578,
0.00200307217108402, 0.00165016998349207, 0.00135877837976224,
0.00111817834219132, 0.000919518759694046, 0.000755491588934199,
0.000620062540625639, 0.000508249477581297, 0.000415939214900695,
0.000339735906332031, 0.000276836029518111, 0.000224925814050528,
0.000182095460069257, 0.000146769499059896, 0.000117648326182113,
9.36606547647744e-05, 7.39240003391358e-05, 5.77123463706654e-05,
4.44292627620613e-05, 3.35069132161542e-05, 2.30948965054054e-05,
1.35498521268301e-05, 6.6480335943903e-06, 3.22817353776792e-06,
1.56754636933548e-06, 7.6117394173847e-07, 3.69613102610182e-07,
1.79477797412926e-07, 8.71511930430619e-08, 4.23189904393637e-08,
2.05492170784783e-08, 9.97819224010812e-09, 4.84507812590914e-09,
2.3525233197402e-09, 1.1421808424864e-09, 5.54459053511967e-10,
2.69071150351853e-10, 1.30491713616467e-10, 6.32000285938001e-11,
3.05246463595271e-11, 1.46586964219514e-11, 6.95585084010912e-12,
3.218335275606e-12, 1.40934851021314e-12, 5.43057275709068e-13,
1.47366028903012e-13, 6.63775977550542e-15)), class = c("tbl_df",
"tbl", "data.frame"), row.names = c(NA, -205L))
plot:
ylim.prim <- c(min(df$y1), max(df$y1))
ylim.sec <- c(min(df$y2), max(df$y2))
b <- diff(ylim.prim)/diff(ylim.sec)
a <- ylim.prim[1] - b*ylim.sec[1]
ggplot(df, aes(x=x)) +
geom_line(aes(y=y1, color = "y1"))+
geom_line(aes(y = 10^(a+ y2*b), color = "y2")) +
scale_y_log10(sec.axis = sec_axis(~ (log(., 10.0) - a)/b,
breaks = 1:6 * 0.5, name = "y2")) +
coord_cartesian(
xlim = c(),
ylim=c(1e-10,3))+
scale_color_brewer(palette = "Set1", name = NULL) +
theme_minimal(base_size = 16) +
theme(axis.line.y.left = element_line(color = "#E41A1C"),
axis.line.y.right = element_line(color = "#377EB8"),
axis.line.x.bottom = element_line("gray"),
axis.text.y.left = element_text(color = "#E41A1C"),
axis.text.y.right = element_text(color = "#377EB8"))
1 answers
solution1
1 ACCPTED 2022-06-20 21:51:52
Showing which line belongs to which axis is simple enough by using colored theme elements and mapping each line to a color aesthetic.
The y1 series is approximately linear if we switch to scale_y_log10 , which is also easy to do.
The difficult part is trying to get the second series mapped to a linear continuous axis when the primary axis is log scaled. To do this, we need to use the value of the second series as an exponent of some small number, chosen so that the scales of the two lines are approximately equal on your plot. We then need to take the log to the same base as part of our transformation in the secondary axis. In your case, 1.03 gives reasonable results:
ggplot(df, aes(x=x)) +
geom_line(aes(y=y1, color = "y1"))+
geom_line(aes(y = 1.05^(a + y2*b), color = "y2")) +
scale_y_log10(sec.axis = sec_axis(~ (log(., 1.03) - a)/b,
breaks = 1:4 * 2.5, name = "y2")) +
scale_color_brewer(palette = "Set1", name = NULL) +
theme_minimal(base_size = 16) +
theme(axis.line.y.left = element_line(color = "#E41A1C"),
axis.line.y.right = element_line(color = "#377EB8"),
axis.line.x.bottom = element_line("gray"),
axis.text.y.left = element_text(color = "#E41A1C"),
axis.text.y.right = element_text(color = "#377EB8"))
You can see that y1 is plotted on a log axis, but y2 is on a normal continuous axis.
As a caveat here, secondary axes are often frowned upon in data visualization, and having one axis on a log scale with the other being non-log is likely to make the plot very difficult for your audience to interpret.
Edit
A more general approach would be to log transform the series on the primary axis and transform its labels too. This allows using a continuous axis rather than log axis:
a <- range(log10(df$y1))
b <- range(df$y2)
ggplot(df, aes(x=x)) +
geom_line(aes(y=(log10(y1) - a[1])/diff(a), color = "y1"))+
geom_line(aes(y = (y2 - b[1])/diff(b) , color = "y2")) +
scale_y_continuous(
labels = ~ 10^(. * diff(a) + a[1]),
breaks = (pretty(log10(df$y1)) - a[1])/diff(a), name = "y1",
sec.axis = sec_axis(~ diff(b) * . + b[1], name = "y2")) +
scale_color_brewer(palette = "Set1", name = NULL) +
theme_minimal(base_size = 16) +
theme(axis.line.y.left = element_line(color = "#E41A1C"),
axis.line.y.right = element_line(color = "#377EB8"),
axis.line.x.bottom = element_line("gray"),
axis.text.y.left = element_text(color = "#E41A1C"),
axis.text.y.right = element_text(color = "#377EB8"))
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.
Related Question
R: custom tick markers for y-axis using log10 scale in ggplot
log10 y-axis labels do not appear on R ggaluvial ggplot ; cannot nudge strata labels to vertical center
Log10 Y-Axis starting from 0
Log10 x-axis on top and y-axis geom_line
How to display p-values above boxplots on exponential (log10) y-axis?
Expand y axis to negative numbers on a log10 scale axis in ggplot2
Making an axis give Natural Log instead of log10 when log = “y” in R plot
ggplot boxplot: log10 y axis AND forced 0-100 range
ggplot plot x-axis transferred with log10 issue
ggplot graph with double y-axis
Related Tags