用ggplot2和直接标签标记点

本文介绍了用ggplot2和直接标签标记点的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！ 问题描述 29岁程序员，3月因学历无情被辞！ 这是我的 按建议绘制通过 解决方案避免重叠的一种方法是（至少在某种程度上）数量由其最接近的点确定。例如，如果一个点的最近邻点直接在它的右边，它的标签将被放置在左边等。 测试中心和正常化变量测试$ yy< $ xx< - （test $ x-min（test $ x））/（max（test $ x） - min（test $ x）） test $ angle for i in 1：nrow（test））{ dx< - test [-i，] $ xx - test [i，] $ xx dy j theta test [i，] $ angle sc test $ nudge.x test $ nudge.y ggplot（aes（x = x，y = y））+ geom_point（aes（color = group））+ geom_text（aes（x = x + nudge.x，y = y + nudge.y，label = ID），size = 3，show.legend = FALSE） 您可以尝试使用缩放参数 sc （它越大，标签离点越远），以避免重叠标签。 （我想可能会发生这样的情况：不同的 sc 可以应用于所有点以避免重叠 - 在这种情况下，您需要更改每个点的缩放参数，可能是通过定义 sc 使用 dx 和 dy ）。 This is a follow-up to my earlier question. While the answer by krlmlr somehow helps to solve my problem, one issue remains - the labels are moved far away from the points such that it gets impossible to see which label corresponds to which point. Does anyone know how to fix this?Test datatest <- structure(list(ID = c(183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 202, 203, 204 ), group = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 10L, 1L, 1L, 11L, 1L, 10L, 10L, 1L, 1L, 1L), .Label = c("a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z"), class = "factor"), x = c(27.4921834914348, 25.9627681619548, 30.4138361188149, 29.7795838507818, 32.33351964819, 31.9669266268744, 35.4433172141507, 37.8161067968601, 43.1590654001956, 44.2747819266045, 46.3829222044342, 42.2074195880057, 45.4532493462957, 48.393281430824, 51.7145681408198, 51.4911582677409, 51.9582538485293, 51.3328966791848, 36.4638478746633, 36.6113033420253, 39.3476493044906), y = c(-40.6667236544384, -39.2640436250506, -40.6403794753022, -40.907139119954, -39.9691483441288, -39.7485517513382, -38.595662907188, -38.2106224386729, -37.6418188651769, -32.7096448721895, -34.1161006958616, -32.2821584775403, -30.8436917254975, -30.3865899667262, -30.3910690518699, -26.1013343566452, -23.8437232732877, -21.5548787351057, 5.50922747751602, 5.64434551903915, 5.01263995541617)), row.names = c(NA, -21L), .Names = c("ID", "group", "x", "y"), class = "data.frame")library(ggplot2)library(directlabels)Plot using geom_text - problem: labels overlapggplot(test, aes(x=x, y=y)) +geom_point(aes(colour=group)) +geom_text(aes(label=ID), show_guide=F)Plot as suggested by krlmlr - problem: labels are far away from pointsggplot(test, aes(x=x, y=y)) +geom_point(aes(colour=group)) +geom_dl(aes(label = ID), method = defaultpf.ggplot("point",,,)) 解决方案 One way to avoid overlapping (to some degree at least) would be to offset each label by an amount which is determined by the closest point to it. So for example if a point's closest neighbouring point is directly to the right of it, its label would be placed to the left, etc. # centre and normalise variablestest$yy <- (test$y - min(test$y)) / (max(test$y) - min(test$y))test$xx <- (test$x - min(test$x)) / (max(test$x) - min(test$x))test$angle <- NAfor (i in 1:nrow(test)) { dx <- test[-i, ]$xx - test[i, ]$xx dy <- test[-i, ]$yy - test[i, ]$yy j <- which.min(dx ^ 2 + dy ^ 2) theta <- atan2((test[-i, ]$yy[j] - test[i, ]$yy), (test[-i, ]$xx[j] - test[i, ]$xx)) test[i, ]$angle <- theta + pi}sc <- 0.5test$nudge.x <- cos(test$angle) * sctest$nudge.y <- sin(test$angle) * scggplot(test, aes(x=x, y=y)) + geom_point(aes(colour=group)) + geom_text(aes(x = x + nudge.x, y = y + nudge.y, label = ID), size = 3, show.legend = FALSE)You can try playing around with the scaling parameter sc (the larger it is, the further away the labels will be from the points) to avoid overlapping labels. (I guess it may happen that not the same sc can be applied to all points to avoid overlaps - in that case you need to change the scaling parameter for each point maybe by defining sc using dx and dy). 这篇关于用ggplot2和直接标签标记点的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！ 上岸，阿里云！