[英]How can I add time dimension in polar coordinates in R?
I'm trying to plot with ggplot2
the track of a bird around the antarctic. 我试图用
ggplot2
绘制一只鸟在南极周围的轨迹。 So far I got a map projected in polar coordinates, I also managed to plot the track points correctly and I almost link them correctly BUT... As the track crosses the international DATE & TIME line , ggplot2
is not able to correctly link the 2 points in either sides of the line.So I'm looking for a way to force ggplot to link the points in continuous manner. 到目前为止,我得到了一个以极坐标投影的地图,我还设法正确绘制了轨迹点,我几乎正确地将它们连接起来......但是当轨道越过国际DATE&TIME线时 ,
ggplot2
无法正确链接2在线的两侧各点。所以我正在寻找一种方法来强制ggplot以连续的方式连接点。
Here's my dataset: 这是我的数据集:
Data =>
ID Date Time A1 Lat. Long.
10 12.9.2008 22:00 1 21.14092 70.98817
10 12.9.2008 22:20 1 21.13031 70.97592
10 12.9.2008 22:40 2 21.13522 70.97853
10 12.9.2008 23:00 1 21.13731 70.97817
10 12.9.2008 23:20 3 21.14197 70.97981
10 12.9.2008 23:40 1 21.14156 70.98158
10 12.9.2008 23:40 1 21.14156 70.98158
10 13.9.2008 00:00 2 21.14150 70.98478
10 13.9.2008 00:20 3 21.14117 70.98803
10 13.9.2008 00:40 1 21.14117 70.98803
10 13.9.2008 01:00 2 21.14117 70.98803
....
ID is the bird's ID. ID是鸟的ID。
UPDATE using Nick K code 使用Nick K代码更新
Here's my original graph without using the time dimension and line 这是我的原始图表,没有使用时间维度和行
south_map <- map_data("world") %>% group_by(group)
set.seed(123)
track_df2 <- new_df2
long_diff <- diff(new_df2$Long)
long_diff[long_diff < -180] <- long_diff[long_diff < -180] + 360
long_diff[long_diff > 180] <- long_diff[long_diff > 180] - 360
track_df2$Longitude <- cumsum(c(new_df2$Long[1], long_diff))
ggplot(track_df2, aes(x = track_df2$Long, y = track_df2$Lat)) +
geom_polygon(aes(group = a3_id), data = south_map, colour = "grey", fill = "gainsboro") +
geom_point(aes(colour = factor(a3_id)), size = 2)
A1 defines what the bird is currently doing. A1定义了鸟目前正在做什么。
You don't seem to actually use the time in your plot, but the issue is the longitudes wrapping around -180/180. 你似乎没有在你的情节中实际使用时间,但问题是经度在-180/180左右。 This can be solved using
coord_map
rather than coord_polar
and ensuring that the longitudes don't wrap around. 这可以使用
coord_map
而不是coord_polar
来解决,并确保经度不会回绕。
Load packages and generate sample data 加载包并生成样本数据
library("ggplot2")
library("dplyr")
south_map <- map_data("world") %>% group_by(group) %>% filter(min(lat) <= -20)
set.seed(123)
track <- data.frame(long = cumsum(c(210,
unlist(lapply(c(1, -1), function(x) {
rnorm(50, x * 4, 4)
})))) %% 360 - 180,
lat = cumsum(c(-50, rnorm(100, 0.4, 2))),
A1 = sample(1:3, 101, replace = TRUE))
Ensure that coordinates don't wrap around: 确保坐标不包围:
track_new <- track
long_diff <- diff(track$long)
long_diff[long_diff < -180] <- long_diff[long_diff < -180] + 360
long_diff[long_diff > 180] <- long_diff[long_diff > 180] - 360
track_new$long <- cumsum(c(track$long[1], long_diff))
Plot using aziequidistant projection. 绘图使用aziequidistant投影。 Note that this assumes the North Pole in the centre, so the latitudes are flipped and then corrected with the scale.
请注意,这假设北极位于中心,因此纬度被翻转然后用比例校正。
ggplot(track_new, aes(x = long, y = -lat)) +
geom_polygon(aes(group = group), data = south_map, colour = "grey", fill = "gainsboro") +
coord_map("azequidistant") +
geom_point(aes(colour = factor(A1)), size = 2) +
geom_path(colour = "grey", size = 1) +
scale_x_continuous(breaks = NULL) +
scale_y_continuous("latitude", breaks = 25 * 0:3, labels = -25 * 0:3)
Final plot: 最终情节:
Just for interest, I thought it would be fun to produce an animation of this image. 只是为了兴趣,我认为制作这个图像的动画会很有趣。 Here's the code to do it:
这是执行此操作的代码:
track_new$alpha <- 1
# Setup longitude labels
long_labels <- data.frame(long = 45 * -3:4, lat = -22.5)
long_labels$label <- long_labels$long
long_labels$label[8] <- "\U00B1 180"
long_labels$angle <- long_labels$long + 67.5 + 180 * (long_labels$long >= 45)
# Set up the basic plot
p <- ggplot(track_new, aes(x = long, y = -lat)) +
geom_polygon(aes(group = group), data = south_map, colour = "grey", fill = "gainsboro") +
coord_map("azequidistant", ylim = c(20, 90)) +
geom_point(aes(colour = A1, alpha = alpha), size = 2) +
geom_path(aes(alpha = alpha), colour = "grey", size = 1) +
scale_x_continuous(breaks = NULL) +
scale_y_continuous("latitude", breaks = 22.5 * 0:3, labels = -22.5 * 0:3) +
scale_alpha_identity(guide = "none") +
geom_text(aes(label = label, angle = angle),
data = long_labels, colour = "dark blue", alpha = 0.5, size = 4)
# Produce the animation
p$data$alpha <- 0
for(i in 1:(nrow(track_new) + 10)) {
p$data$alpha <- pmax(p$data$alpha - 0.1, 0)
if (i <= nrow(track_new)) {
p$data$alpha[i] <- 1
}
png(file.path("BirdPlots", sprintf("BirdPlot%03d.png", i)), width = 1024, height = 1024, res = 100)
print(p)
dev.off()
if (!(i %% 5)) cat(i, "\n")
}
# This needs ImageMagick in the system path. For non-Windows systems, you
# might be better using system rather than shell
shell(paste("convert", file.path("BirdPlots", "BirdPlot*.png"),
file.path("BirdPlots", "BirdPlotAnimation.gif")))
And here's the result: 这是结果:
EDIT Corrected version of ayush's code 编辑修正了ayush代码的版本
track_df2 <- new_df2
long_diff <- diff(new_df2$Longitude)
long_diff[long_diff < -180] <- long_diff[long_diff < -180] + 360
long_diff[long_diff > 180] <- long_diff[long_diff > 180] - 360
track_df2$Longitude <- cumsum(c(new_df2$Longitude[1], long_diff))
track_df2$a3_id <- factor(track_df2$a3_id)
ggplot(track_df2, aes(x = Longitude, y = -Latitude)) +
coord_map("azequidistant", ylim = c(20, 90)) +
geom_point(aes(colour = a3_id, alpha = alpha), size = 2) +
geom_path(aes(alpha = alpha), colour = "grey", size = 1) +
scale_x_continuous(breaks = NULL) +
scale_y_continuous(breaks = 22.5 * 0:3, labels = -22.5 * 0:3) +
scale_alpha_identity(guide = "none")
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.