在 ggplot 中绘制带有图例的时间序列的问题（股票价格）答案

【问题标题】：Issues with plotting a time series in ggplot, with legends (Stock prices)在 ggplot 中绘制带有图例的时间序列的问题（股票价格）
【发布时间】：2020-12-18 19:38:33
【问题描述】：

我正在尝试用指标绘制股票价格，但是我不知道如何在图中添加标签。我想我无法理解 ggplot 如何处理时间序列。我已经看到其他人使用所有数据创建新列，但我不知道这将如何应用在这里。我最初想绘制收盘价以及 50 天和 200 天移动平均线，以及相应的标签。

这里是没有标签的数据和绘图。也欢迎任何有关了解 ggplot 的资源。

structure(c(3.320236, 3.300589, 3.483955, 3.588736, 3.418468, 
3.497053, 3.549443, 3.582187, 3.601834, 3.588736, 1803600, 2480600, 
2608100, 2315800, 3706400, 2648200, 1851400, 1297000, 2230600, 
1667900, 4.04544857999999, 4.01964633999999, 4.00091679999999, 
3.98559263999999, 3.96987555999999, 3.95442041999999, 3.94407331999999, 
3.93791745999999, 3.92586769999999, 3.91892597999999, 4.80337260999993, 
4.79541582999993, 4.78650947499993, 4.77812702499993, 4.76981006499993, 
4.76195151999993, 4.75271772499993, 4.74482643499993, 4.73775374499993, 
4.72956775999993, 169569500, 167088900, 169697000, 172012800, 
168306400, 170954600, 172806000, 174103000, 176333600, 174665700, 
3.32023599999963, 3.30058899999973, 3.48395499999974, 3.58873599999971, 
3.41846799999982, 3.49705299999975, 3.54944299999964, 3.58218699999948, 
3.6018339999997, 3.5887359999996), class = c("xts", "zoo"), index = structure(c(1546387200, 
1546473600, 1546560000, 1546819200, 1546905600, 1546992000, 1547078400, 
1547164800, 1547424000, 1547510400), tzone = "UTC", tclass = "Date"), .Dim = c(10L, 
6L), .Dimnames = list(NULL, c("SPWR.Close", "SPWR.Volume", "SMA", 
"SMA.1", "obv", "VWAP")))

ggplot(data = df, aes(x = Date, y = Price)) + 
  geom_line(aes(y = SPWR.Close)) +
  geom_line(aes(y = SMA), col = "red") +
  geom_line(aes(y = SMA.1), col = "blue")

【问题讨论】：

请正确添加dput(yourdata)，包含的有错别字！
实际上实现了一个不错的时间序列方法，试试plot(df)（必须加载library(xts)）。

标签： r ggplot2 charts time-series stock

【解决方案1】：

一种将 TS 转换为数据框并将行名转换为新列的方法。然后重新排列数据以简洁地绘制它

# needed libraries
library(ggplot2)
library(dplyr)
library(tidyr)
# your dummy data
df <- structure(c(3.320236, 3.300589, 3.483955, 3.588736, 3.418468, 
              3.497053, 3.549443, 3.582187, 3.601834, 3.588736, 1803600, 2480600, 
              2608100, 2315800, 3706400, 2648200, 1851400, 1297000, 2230600, 
              1667900, 4.04544857999999, 4.01964633999999, 4.00091679999999, 
              3.98559263999999, 3.96987555999999, 3.95442041999999, 3.94407331999999, 
              3.93791745999999, 3.92586769999999, 3.91892597999999, 4.80337260999993, 
              4.79541582999993, 4.78650947499993, 4.77812702499993, 4.76981006499993, 
              4.76195151999993, 4.75271772499993, 4.74482643499993, 4.73775374499993, 
              4.72956775999993, 169569500, 167088900, 169697000, 172012800, 
              168306400, 170954600, 172806000, 174103000, 176333600, 174665700, 
              3.32023599999963, 3.30058899999973, 3.48395499999974, 3.58873599999971, 
              3.41846799999982, 3.49705299999975, 3.54944299999964, 3.58218699999948, 
              3.6018339999997, 3.5887359999996), class = c("xts", "zoo"), index = structure(c(1546387200, 1546473600, 1546560000, 1546819200, 1546905600, 1546992000, 1547078400, 1547164800, 1547424000, 1547510400), tzone = "UTC", tclass = "Date"), .Dim = c(10L, 6L), .Dimnames = list(NULL, c("SPWR.Close", "SPWR.Volume", "SMA", "SMA.1", "obv", "VWAP"))) 

# convert to df
df1 <- as.data.frame(df)
# rownames (dates) to a new column
df1$DATE <- rownames(df1)

df1 %>% 
  # get data in a better format to print
  tidyr::pivot_longer(-DATE, names_to = "Variable", values_to = "Values") %>% 
  # fitler the variable you want to print
  dplyr::filter(Variable %in% c("SPWR.Close", "SMA", "SMA.1")) %>% 
  # plot using the group feature of ggplot2
  ggplot2::ggplot(aes(x = DATE, y = Values, group = Variable, color = Variable)) +
  ggplot2::geom_line()

【讨论】：

是的，这很好用，但是如果我想使用geom_point 函数将其中一个图作为散点图怎么办？或者假设我想将 SPWR.volume 绘制为底部的条形图？
最简单的方法是不过滤任何变量，然后在 geom_line() 调用之后使用 facet_wrap(~Variable, scales = "free")（它们之间有一个 +） - 这样您将有 4 个单独的折线图，每个变量一个，每个都有自己的比例。其他的都比较复杂

【解决方案2】：

您有一个 xts/zoo 系列，但问题中缺少相关的库调用。 zoo 有自己的经典图形 (plot.zoo)、格 (xyplot.zoo) 和 ggplot2 (autoplot.zoo) 的绘图方法。

我们使用收盘点、交易量直方图柱和所有其他列使用 3 个面板来说明每一个，第二个面板用于交易量，第三个用于 OBV，所有其他列都绘制在第一个面板中。

1) plot.zoo 使用 plot.zoo（我们在末尾显示 autoplot.zoo）参数 screen、col 和 type 每列有一个元素，通过以下方式指定面板编号（或标签） screen=, color via col= and type (p = point, l = line, h = histogram) via type= for that column。在图例中，我们在第一个参数中指定位置，在第二个参数中指定图例文本，通过 col= 指定颜色，通过 pch= 指定绘制为点的列的点类型（20 = 填充点），绘制为线条的列的线类型（1 =实线）通过 lty=，线宽通过 lwd=，点大小通过 pt.cex=，文本大小通过 cex=，剪切（TRUE = 抑制剪切）通过 xpd= 和边框（n = 图例周围无边框）通过 bty=.

library(xts)

plot(transform(as.zoo(x), 
       SPWR.Volume = SPWR.Volume / 1000,
       obv  = obv / 1000000),
  main = "SPWR",
  ylab = c("Price", "Volume", "OBV"),
  screen = c(1, 2, 1, 1, 3, 1), 
  col = c("black", "blue", "green", "red", "black", "purple"),
  type = c("p", "h", "l", "l", "l", "l"),
  lwd = 2,
  heights = c(4, 1, 1)
)
legend("topleft", c("SMA", "SMA.1", "VWAP", "Close"), 
  col = c("green", "red", "purple", "black"), 
  pch = c(NA, NA, NA, 20), lty = c(1, 1, 1, NA),
  lwd = 2, pt.cex = 1.2, cex = 0.7, xpd = TRUE, bty = "n")

2) xyplot.zoo xyplot.zoo 使用点阵图形。该调用有点类似于 plot.zoo。

library(lattice)
library(xts)

xyplot(transform(as.zoo(x),
                 SPWR.Volume = SPWR.Volume / 1000,
                 obv  = obv / 1000000),
  main = "SPWR",
  ylab = c("Price", "Volume", "OBV"),
  screen = c("Price", "Volume", "Price", "Price", "OBV", "Price"),
  col = c("black", "blue", "green", "red", "black", "purple"),
  type = list("p", "h", "l", "l", "l", "l"),
  lwd = 2,
  par.settings = list(layout.heights = list(panel = c(4, 1, 1)/6)),
  key = list(corner = c(0.05, 0.9),
    lines = list(col = c("green", "red", "purple")),
    text = list(c("SMA", "SMA.1", "VWAP"), cex = 0.7)
  )
)

3) autoplot.zoo ggplot2 相当复杂，但我们可以通过创建 3 个单独的图然后使用 cowplot 组合它们来实现。

library(cowplot)
library(ggplot2)
library(xts)

xx <- transform(as.zoo(x), 
  SPWR.Volume = SPWR.Volume / 1000,
  obv  = obv / 1000000)
p1 <- autoplot(xx$SPWR.Close, geom = "point") + 
  geom_line(aes(y = VWAP, col = "A"), x) +
  geom_line(aes(y = SMA, col = "B"), x) + 
  geom_line(aes(y = SMA.1, col = "C"), x) + 
  scale_color_manual(name = "Legend", 
    values = c(A = "purple", B = "green", C = "red"), 
    labels = c(A = "VWAP", B = "SMA", C = "SMA.1")) +
  theme(legend.position = c(0.1, 0.73), 
    plot.margin = margin(1, 1, -0.5, 1, "cm"),
    axis.text.x = element_blank(),
    axis.ticks.x = element_blank()) +
  ggtitle("SPWR") +
  xlab("") + 
  ylab("Price")
p2 <- autoplot(xx$SPWR.Volume, fill = "Volume", geom = "col") +
  scale_fill_manual(values = "blue") + 
  guides(fill = FALSE) +
  theme(plot.margin = margin(-0.5, 1, -0.5, 1, "cm")) +
  xlab("") + 
  ylab("Volume")
p3 <- autoplot(xx$obv) +
  theme(plot.margin = margin(-0.5, 1, 1, 1, "cm")) +
  ylab("OBV")
plot_grid(p1, p2, p3, align = "v", rel_heights = c(4, 1, 1), ncol = 1)

【讨论】：