热门标签 | HotTags
当前位置:  开发笔记 > 编程语言 > 正文

在单个图表中实现饼图与条形图的精准对齐

在看到有关如何在 ggplot2 中从经济学家那里重新创建此图的问题后,我决定自己从头开始尝试(因为没有提供代码或数据),因为我发现这很有趣。这是我迄今为止设法做的事情:我能够相对轻松地做到这一点。但

在看到有关如何在 ggplot2 中从经济学家那里重新创建此图的问题后,我决定自己从头开始尝试(因为没有提供代码或数据),因为我发现这很有趣。

这是我迄今为止设法做的事情:

我能够相对轻松地做到这一点。但是,我正在努力放置饼图。因为 ggplot 使用笛卡尔坐标来制作饼图,所以我不能在同一个图表上有条形图和饼图。所以我发现geom_arc_bar()from ggforce,它确实允许笛卡尔坐标系上的馅饼。但是,问题出在coord_fixed(). 我可以让馅饼对齐,但没有coord_fixed(). 但是,对于coord_fixed(),我无法使图表与经济学人图表的高度相匹配。没有coord_fixed()我可以,但馅饼是椭圆形而不是圆形。见下文:

coord_fixed()

没有coord_fixed()

我尝试过的另一个选择是分别制作一系列饼图,然后将这些图组合在一起。但是,我努力使情节与gridExtra其他替代方案保持一致。我确实结合了油漆。显然这是有效的,但不是程序化的。我需要一个 100% 基于 R 的解决方案。

我在油漆中粘贴来自 R 的单独图像的解决方案:

有人有解决这个问题的方法吗?我认为这是一个有趣的问题,我提供了一个起点。我愿意接受任何建议,也可以随意提出完全不同的方法,因为我承认我的不是最好的。谢谢!

代码:

# packages
library(data.table)
library(dplyr)
library(forcats)
library(ggplot2)
library(ggforce)
library(ggnewscale)
library(ggtext)
library(showtext)
library(stringr)
# data
global <- fread("Sector,ROE,Share,Status
Technology,14.2,10,Local
Technology,19,90,Multinational
Other consumer,16.5,77,Multinational
Other consumer,20.5,23,Local
Industrial,13,70,Multinational
Industrial,18,30,Local
Cyclical consumer,12,77,Multinational
Cyclical consumer,21,23,Local
Utilities,6,88,Local
Utilities,11,12,Multinational
All sectors,10,50,Local
All sectors,10.2,50,Multinational
Financial,6,27,Multinational
Financial,10.5,73,Local
Diversified,4.9,21,Local
Diversified,5,79,Multinational
Basic materials,4,82,Multinational
Basic materials,9,18,Local
Media & communications,3,76,Multinational
Media & communications,14,24,Local
Energy,-1,40,Local
Energy,1,60,Multinational
")
equity <- global %>%
group_by(Sector) %>%
mutate(xend = ifelse(min(ROE) > 0, 0, min(ROE)))
equity$Sector <- factor(equity$Sector, levels= rev(c("Technology", "Other consumer",
"Industrial", "Cyclical consumer",
"Utilities", "All sectors", "Financial",
"Diversified", "Basic materials",
"Media & communications", "Energy")))
equity$Status <- factor(equity$Status, levels = c("Multinational", "Local"))
# fonts
font_add_google("Montserrat", "Montserrat")
font_add_google("Roboto", "Roboto")
# scaling text for high res image
img_scale <- 5.5
# graph
showtext_auto() # for montserrat font to show
economist <- ggplot(equity)+
geom_vline(aes(xintercept = -2.5, color = "+-"), show.legend = FALSE)+
geom_vline(aes(xintercept = 2.5, color = "+-"), show.legend = FALSE)+
geom_segment(aes(x = ROE, xend = xend, y = Sector, yend = Sector, color = "line"),
show.legend = FALSE, size = 2)+
geom_tile(aes(x = ROE, y = Sector, width = 1, height = 0.5, fill = Status),
size = 0.5)+
geom_vline(aes(xintercept = 0, color = "x-axis"), show.legend = FALSE)+
scale_fill_manual("", values = c("Local" = "#ea5f47", "Multinational" = "#0a5268"))+
scale_color_manual(values = c("x-axis" = "red", "+-" = "#cddee6", "line" = "#a8adb3"))+
scale_x_continuous(position = "top", limits = c(-5, 25),
breaks = c(-5, -2.5, 0, 2.5, 5,10,15,20,25),
labels = c(5, "-", 0, "+", 5,10,15,20,25),
minor_breaks = c(-2.5, 2.5)
)+
scale_y_discrete(labels = function(x) str_replace_all(x, "& c" , "&nc"))+
#width = 40))+
labs(x = "", y = "", caption = c("Sources: Bloomberg;",
"The Economist",
"*Top 500 global companies"))+
ggtitle("The price of being global",
subtitle = "Return on equity*, latest 12 months, %")+
theme(legend.position = "top",
legend.direction = "vertical",
legend.justification = -1.25,
legend.key.size = unit(0.18, "cm"),
legend.key.height = unit(0.1, "cm"),
legend.background = element_rect("#cddee6"),
legend.text = element_text("Montserrat", size = 9 * img_scale),
plot.background = element_rect("#cddee6"),
plot.margin = margin(t = 10, r = 10, b = 20, l = 10),
panel.background = element_rect("#cddee6"),
panel.grid.major.y = element_blank(),
panel.grid.minor.y = element_blank(),
panel.grid.minor.x = element_blank(),
axis.ticks = element_blank(),
axis.text = element_text(family = "Montserrat", size = 9 * img_scale,
colour = "black"),
axis.text.y = element_text(hjust = 0, lineheight = 0.15,
face = c(rep("plain",5), "bold.italic", rep("plain",5))
),
#axis.text.x = element_text(family = "Montserrat", size = 9*img_scale,)
plot.title = element_text(family = "Montserrat", size = 12 * img_scale,
face = "bold",
hjust = -34.12),
text = element_text(family = "Montserrat"),
plot.subtitle = element_markdown(family = "Montserrat", size = 9 * img_scale,
hjust = 7.5),
plot.caption = element_markdown(size = 9*img_scale,
face = c("plain", "italic", "plain"),
hjust = c(-1.35, -1.85, -2.05),
vjust = c(0,0.75,0)))
# only way to get google fonts on plot (R device does not show them)
png("bar.png", height = 480*8, width = 250*8, res = 72*8) # increased resolution (dpi)
economist
dev.off()
# piechart
pies <- equity %>%
mutate(Sector = fct_rev(Sector)) %>%
ggplot(aes(x = "", y = Share, fill = Status, width = 0.15)) +
geom_bar(stat = "identity", position = position_fill(), show.legend = FALSE, size = 0.1) +
# geom_text(aes(label = Cnt), position = position_fill(vjust = 0.5)) +
coord_polar(theta = "y", direction = -1) +
facet_wrap(~ Sector, dir = "v", ncol = 1) +
scale_fill_manual("", values = c("Local" = "#93b7c7", "Multinational" = "#08526b"))+
#theme_void()+
theme(panel.spacing = unit(-0.35, "lines"),
plot.background = element_rect("#cddee6"),
panel.background = element_rect("transparent"),
strip.text = element_blank(),
axis.title.x = element_blank(),
axis.title.y = element_blank(),
legend.position='none',
axis.ticks = element_blank(),
axis.text = element_blank(),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank())
# guides(fill=guide_legend(nrow=2, byrow=TRUE))
png("pie_chart.png", height = 350*8, width = 51*8, res = 72*8)
pies
dev.off()
# geom_bar_arc (ggforce) with coord_fixed - cannot match height but pies are circular
eco_circle_pies <- ggplot(equity)+
geom_vline(aes(xintercept = -2.5, color = "+-"), show.legend = FALSE)+
geom_vline(aes(xintercept = 2.5, color = "+-"), show.legend = FALSE)+
geom_segment(aes(x = ROE, xend = xend, y = Sector, yend = Sector, color = "line"),
show.legend = FALSE, size = 1)+
scale_fill_manual("", values = c("Local" = "#ea5f47", "Multinational" = "#0a5268"))+
geom_tile(aes(x = ROE, y = Sector, width = 1, height = 0.5, fill = Status),
size = 0.5, show.legend = TRUE)+
geom_vline(aes(xintercept = 0, color = "x-axis"), show.legend = FALSE)+
new_scale_fill()+
geom_arc_bar(aes(x0 = 27, y0 = as.numeric(equity$Sector), r0 = 0, r = 0.45,
amount = Share,
fill = Status),
stat = 'pie',
color = "transparent",
show.legend = FALSE)+
coord_fixed()+
scale_fill_manual("", values = c("Local" = "#93b7c7", "Multinational" = "#08526b"))+
scale_color_manual(values = c("x-axis" = "red", "+-" = "#cddee6", "line" = "#a8adb3"))+
scale_x_continuous(position = "top", limits = c(-5, 30),
breaks = c(-5, -2.5, 0, 2.5, 5,10,15,20,25),
labels = c(5, "-", 0, "+", 5,10,15,20,25),
minor_breaks = c(-2.5, 2.5)
)+
scale_y_discrete(labels = function(x) str_replace_all(x, "& c" , "&nc"))+
# below is to get * superscript
labs(x = "", y = "", caption = c("Sources: Bloomberg;",
"The Economist",
"*Top 500 global companies"))+ # this is to get
ggtitle("The price of being global",
subtitle = "Return on equity*, latest 12 months, %")+
guides(color = FALSE)+
theme(legend.position = "top",
legend.direction = "vertical",
# legend.justification = -0.9,
legend.key.size = unit(0.18, "cm"),
legend.key.height = unit(0.1, "cm"),
legend.background = element_rect("#cddee6"),
legend.text = element_text("Montserrat", size = 9 * img_scale),
plot.background = element_rect("#cddee6"),
# plot.margin = margin(t = -80, r = 10, b = -20, l = 10),
panel.background = element_rect("#cddee6"),
panel.grid.major.y = element_blank(),
panel.grid.minor.y = element_blank(),
panel.grid.minor.x = element_blank(),
axis.ticks = element_blank(),
axis.text = element_text(family = "Montserrat", size = 9 * img_scale,
colour = "black"),
axis.text.y = element_text(hjust = 0, lineheight = 0.15),
#axis.text.x = element_text(family = "Montserrat", size = 9*img_scale,)
plot.title = element_text(family = "Montserrat", size = 12 * img_scale,
hjust = -2.12),
plot.subtitle = element_markdown(family = "Montserrat", size = 9 * img_scale,
hjust = -5.75),
plot.caption = element_markdown(size = 9*img_scale,
face = c("plain", "italic", "plain"),
#hjust = c(-.9, -1.22, -1.95),
#vjust = c(0,0.75,0)))
))
png("eco_circle_pies.png", height = 220*8, width = 420*8, res = 72*8)
eco_circle_pies
dev.off()
# geom_bar_arc (ggforce) without coord_fixed - matches height, but pies are oval
eco_oval_pie <- ggplot(equity)+
geom_vline(aes(xintercept = -2.5, color = "+-"), show.legend = FALSE)+
geom_vline(aes(xintercept = 2.5, color = "+-"), show.legend = FALSE)+
geom_segment(aes(x = ROE, xend = xend, y = Sector, yend = Sector, color = "line"),
show.legend = FALSE, size = 1)+
scale_fill_manual("", values = c("Local" = "#ea5f47", "Multinational" = "#0a5268"))+
geom_tile(aes(x = ROE, y = Sector, width = 1, height = 0.5, fill = Status),
size = 0.5, show.legend = TRUE)+
geom_vline(aes(xintercept = 0, color = "x-axis"), show.legend = FALSE)+
new_scale_fill()+
geom_arc_bar(aes(x0 = 27, y0 = as.numeric(equity$Sector), r0 = 0, r = 0.45,
amount = Share,
fill = Status),
stat = 'pie',
color = "transparent",
show.legend = FALSE)+
# coord_fixed()+
scale_fill_manual("", values = c("Local" = "#93b7c7", "Multinational" = "#08526b"))+
scale_color_manual(values = c("x-axis" = "red", "+-" = "#cddee6", "line" = "#a8adb3"))+
scale_x_continuous(position = "top", limits = c(-5, 30),
breaks = c(-5, -2.5, 0, 2.5, 5,10,15,20,25),
labels = c(5, "-", 0, "+", 5,10,15,20,25),
minor_breaks = c(-2.5, 2.5)
)+
scale_y_discrete(labels = function(x) str_replace_all(x, "& c" , "&nc"))+
#width = 40))+
labs(x = "", y = "", caption = c("Sources: Bloomberg;",
"The Economist",
"*Top 500 global companies"))+
ggtitle("The price of being global",
subtitle = "Return on equity*, latest 12 months, %")+
guides(color = FALSE)+
theme(legend.position = "top",
legend.direction = "vertical",
legend.justification = -1.1,
legend.key.size = unit(0.18, "cm"),
legend.key.height = unit(0.1, "cm"),
legend.background = element_rect("#cddee6"),
legend.text = element_text("Montserrat", size = 9 * img_scale),
plot.background = element_rect("#cddee6"),
# plot.margin = margin(t = -80, r = 10, b = -20, l = 10),
panel.background = element_rect("#cddee6"),
panel.grid.major.y = element_blank(),
panel.grid.minor.y = element_blank(),
panel.grid.minor.x = element_blank(),
axis.ticks = element_blank(),
axis.text = element_text(family = "Montserrat", size = 9 * img_scale,
colour = "black"),
axis.text.y = element_text(hjust = 0, lineheight = 0.15),
text = element_text(family = "Montserrat"),
plot.title = element_text(family = "Montserrat", size = 12 * img_scale,
face = "bold",
hjust = -7.05),
plot.subtitle = element_markdown(family = "Montserrat", size = 9 * img_scale,
hjust = 53.75),
plot.caption = element_markdown(size = 9*img_scale,
face = c("plain", "italic", "plain"),
hjust = c(-1.15, -1.58, -1.95),
vjust = c(0.5,1.15,0.5)))
png("eco_oval_pies.png", height = 480*8, width = 250*8, res = 72*8)
eco_oval_pie
dev.off()

回答


确实是一个有趣的问题。在我看来,获得所需结果的最简单方法是创建两个单独的图并使用美妙的patchwork包将它们粘合在一起:

注意:为了关注主要问题并使代码更简洁,我放弃了所有或大部分的主题调整、ggtext样式、自定义字体……。相反,我依靠ggthemes::theme_economist接近经济学家的样子。

# packages
library(data.table)
library(dplyr)
library(stringr)
library(forcats)
library(ggplot2)
library(patchwork)
library(ggthemes)
bars <-ggplot(equity)+
geom_vline(aes(xintercept = -2.5, color = "+-"), show.legend = FALSE)+
geom_vline(aes(xintercept = 2.5, color = "+-"), show.legend = FALSE)+
geom_segment(aes(x = ROE, xend = xend, y = Sector, yend = Sector, color = "line"),
show.legend = FALSE, size = 2)+
geom_tile(aes(x = ROE, y = Sector, width = 1, height = 0.5, fill = Status),
size = 0.5)+
geom_vline(aes(xintercept = 0, color = "x-axis"), show.legend = FALSE)+
scale_fill_manual("", values = c("Local" = "#ea5f47", "Multinational" = "#0a5268"))+
scale_color_manual(values = c("x-axis" = "red", "+-" = "#cddee6", "line" = "#a8adb3"))+
scale_x_continuous(position = "top", limits = c(-5, 25),
breaks = c(-5, -2.5, 0, 2.5, 5,10,15,20,25),
labels = c(5, "-", 0, "+", 5,10,15,20,25),
minor_breaks = c(-2.5, 2.5)
)+
scale_y_discrete(labels = function(x) str_replace_all(x, "& c" , "&nc"))+
labs(x = "", y = "") +
ggthemes::theme_economist() +
theme(legend.position = "top", legend.justification = "left")
pies <- equity %>%
mutate(Sector = fct_rev(Sector)) %>%
ggplot(aes(x = "", y = Share, fill = Status, width = 0.15)) +
geom_bar(stat = "identity", position = position_fill(), show.legend = FALSE, size = 0.1) +
coord_polar(theta = "y", direction = -1) +
facet_wrap(~ Sector, dir = "v", ncol = 1) +
scale_fill_manual("", values = c("Local" = "#93b7c7", "Multinational" = "#08526b")) +
labs(x = NULL, y = NULL) +
ggthemes::theme_economist() +
theme(strip.text = element_blank(), panel.spacing.y = unit(0, "pt"),
axis.text = element_blank(), , axis.ticks = element_blank(), axis.line = element_blank(),
panel.grid.major = element_blank())
bars + pies +
plot_layout(widths= c(5, 1)) +
plot_annotation(caption = c("Sources: Bloomberg;",
"The Economist", "Top 500 global companies"),
title = "The price of being global",
subtitle = "Return on equity, latest 12 months, %",
theme = theme_economist())






推荐阅读
  • 本文详细探讨了 Django 的 ORM(对象关系映射)机制,重点介绍了其如何通过 Python 元类技术实现数据库表与 Python 类的映射。此外,文章还分析了 Django 中各种字段类型的继承结构及其与数据库数据类型的对应关系。 ... [详细]
  • 本题探讨了在一个有向图中,如何根据特定规则将城市划分为若干个区域,使得每个区域内的城市之间能够相互到达,并且划分的区域数量最少。题目提供了时间限制和内存限制,要求在给定的城市和道路信息下,计算出最少需要划分的区域数量。 ... [详细]
  • 20100423:Fixes:更新批处理,以兼容WIN7。第一次系统地玩QT,于是诞生了此预备式:【QT版本4.6.0&#x ... [详细]
  • 本文详细介绍了Python中函数的基本概念,包括函数的定义与调用、文档注释、参数传递(形参与实参)、返回值以及函数嵌套。通过具体示例和解释,帮助读者掌握函数在编程中的应用。 ... [详细]
  • 在现代Web应用中,当用户滚动到页面底部时,自动加载更多内容的功能变得越来越普遍。这种无刷新加载技术不仅提升了用户体验,还优化了页面性能。本文将探讨如何实现这一功能,并介绍一些实际应用案例。 ... [详细]
  • 本文详细介绍了 org.apache.commons.io.IOCase 类中的 checkCompareTo() 方法,通过多个代码示例展示其在不同场景下的使用方法。 ... [详细]
  • 本文探讨了如何在iOS开发环境中,特别是在Xcode 6.1中,设置和应用自定义文本样式。我们将详细介绍实现方法,并提供一些实用的技巧。 ... [详细]
  • 本文详细介绍了C语言的起源、发展及其标准化过程,涵盖了从早期的BCPL和B语言到现代C语言的演变,并探讨了其在操作系统和跨平台编程中的重要地位。 ... [详细]
  • This pull request introduces the ability to provide comprehensive paragraph configurations directly within the Create Note and Create Paragraph REST endpoints, reducing the need for additional configuration calls. ... [详细]
  • 本文介绍如何在SQL Server中创建动态SQL存储过程,并提供详细的代码实例和解释。通过这种方式,可以更灵活地处理查询条件和参数。 ... [详细]
  • 中科院学位论文排版指南
    随着毕业季的到来,许多即将毕业的学生开始撰写学位论文。本文介绍了使用LaTeX排版学位论文的方法,特别是针对中国科学院大学研究生学位论文撰写规范指导意见的最新要求。LaTeX以其精确的控制和美观的排版效果成为许多学者的首选。 ... [详细]
  • 丽江客栈选择问题
    本文介绍了一道经典的算法题,题目涉及在丽江河边的n家特色客栈中选择住宿方案。两位游客希望住在色调相同的两家客栈,并在晚上选择一家最低消费不超过p元的咖啡店小聚。我们将详细探讨如何计算满足条件的住宿方案总数。 ... [详细]
  • 2018-2019学年第六周《Java数据结构与算法》学习总结
    本文总结了2018-2019学年第六周在《Java数据结构与算法》课程中的学习内容,重点介绍了非线性数据结构——树的相关知识及其应用。 ... [详细]
  • 本文介绍了 Winter-1-C A + B II 问题的详细解题思路和测试数据。该问题要求计算两个大整数的和,并输出结果。我们将深入探讨如何处理大整数运算,确保在给定的时间和内存限制下正确求解。 ... [详细]
  • PHP 过滤器详解
    本文深入探讨了 PHP 中的过滤器机制,包括常见的 $_SERVER 变量、filter_has_var() 函数、filter_id() 函数、filter_input() 函数及其数组形式、filter_list() 函数以及 filter_var() 和其数组形式。同时,详细介绍了各种过滤器的用途和用法。 ... [详细]
author-avatar
龙蛇虫
这个家伙很懒,什么也没留下!
PHP1.CN | 中国最专业的PHP中文社区 | DevBox开发工具箱 | json解析格式化 |PHP资讯 | PHP教程 | 数据库技术 | 服务器技术 | 前端开发技术 | PHP框架 | 开发工具 | 在线工具
Copyright © 1998 - 2020 PHP1.CN. All Rights Reserved | 京公网安备 11010802041100号 | 京ICP备19059560号-4 | PHP1.CN 第一PHP社区 版权所有