一般性评论

Question

我正在制作带有一些跟踪数据的2d密度图。

我的目标是为每个实验单元绘制

我的df看起来像这样：

head(df)
  Caja  x  y          t
1    4  0  0 0:00:00:00
2    4 -2 10 0:00:00:14
3    4  3 26 0:00:00:28
4    4  7 54 0:00:00:42
5    4  9 75 0:00:00:57
6    4  9 92 0:00:00:71

第一列（Caja）是我用于分面的因素（因为每个级别是不同的实验单位）。 x和y是我感兴趣的变量，t就是衡量标准的时间。

由于df是一个很长的data.frane，我稍后会使用caja4我提供它

dput(caja4)
structure(list(Caja = c(4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), x = c(0L, -2L, 
3L, 7L, 9L, 9L, 11L, 6L, -12L, -23L, -33L, -32L, -36L, -28L, 
-15L, -2L, 2L, 6L, 6L, 4L, 4L, 2L, 1L, 0L, -1L, 2L, 2L, 4L, 6L, 
11L, 9L, 11L, 13L, 12L, 0L, 3L, 10L, 5L, -14L, -10L, 5L, 10L, 
12L, 12L, 11L, 11L, 9L, 9L, 11L, 9L, 10L, 11L, 12L, 7L, 1L, -6L, 
-11L, -18L, -25L, -17L, -7L, -2L, -5L, -4L, -24L, -29L, -37L, 
-39L, -41L, -41L, -41L, -41L, -41L, -41L, -41L, -39L, -39L, -39L, 
-39L, -39L, -39L, -39L, -39L, -39L, -28L, -17L, -11L, -6L, -9L, 
-16L, -27L, -37L, -38L, -42L, -42L, -42L, -42L, -42L, -42L, -42L, 
-42L, -42L, -42L, -36L, -28L, -17L, -6L, 1L, 3L, 3L, 2L, 2L, 
2L, 2L, -2L, 1L, 0L, 0L, 2L, 4L, 5L, 4L, 2L, 3L, -1L, -1L, -9L, 
-12L, -30L, -38L, -37L, -36L, -42L, -42L, -42L, -42L, -42L, -43L, 
-43L, -43L, -36L, -35L, -29L, -26L, -29L, -37L, -38L, -38L, -38L, 
-37L, -31L, -18L, 5L, 7L, 7L, 8L, 7L, 4L, 7L, 2L, 5L, 10L, 13L, 
12L, -13L, -32L, -41L, -31L, -14L, 0L, 10L, 2L, -8L, -29L, -37L, 
-37L, -37L, -37L, -37L, -37L, -37L, -28L, -14L, -3L, 4L, 3L, 
1L, -7L, -15L, -11L, -11L, -15L, -16L, -27L, -37L, -43L, -43L, 
-43L, -43L, -43L, -43L, -43L, -24L, -1L, 0L, 3L, 4L, 3L, 2L, 
1L, 3L, 4L, -9L, -31L, -45L, -45L, -34L, -32L, -4L, -12L, -8L, 
-6L, -17L, -31L, -36L, -42L, -37L, -39L, -36L, -36L, -36L, -37L, 
-37L, -36L, -37L, -32L, -19L, -5L, 2L, 7L, 9L, 9L, 7L, 10L, 9L, 
10L, 11L, 11L, 11L, 8L, 8L, 10L, 12L, -1L, -14L, -32L, -17L, 
-2L, -14L, -26L, -44L, -44L, -44L, -44L, -44L, -44L, -45L, -45L, 
-33L, -21L, -12L, -11L, -5L, 2L, 5L, 4L, 2L, 6L, 6L, 6L, 6L, 
6L, 6L, 7L, -7L, -29L, -27L, -10L, -2L, 5L, 7L, 7L, 8L, 13L), 
    y = c(0L, 10L, 26L, 54L, 75L, 92L, 113L, 116L, 117L, 114L, 
    119L, 115L, 114L, 108L, 99L, 90L, 76L, 57L, 37L, 21L, 12L, 
    7L, 8L, 2L, -2L, 5L, 7L, 17L, 37L, 52L, 67L, 86L, 100L, 118L, 
    122L, 125L, 122L, 122L, 129L, 129L, 126L, 124L, 125L, 123L, 
    106L, 92L, 70L, 49L, 27L, 17L, 5L, 6L, 10L, 7L, 3L, -3L, 
    -3L, -6L, -8L, -13L, -13L, -14L, -13L, -13L, -5L, 1L, 3L, 
    7L, 8L, 9L, 10L, 25L, 42L, 44L, 57L, 87L, 101L, 107L, 110L, 
    102L, 96L, 94L, 81L, 68L, 57L, 50L, 61L, 76L, 96L, 97L, 96L, 
    98L, 99L, 95L, 79L, 64L, 53L, 46L, 31L, 12L, 4L, -7L, -16L, 
    -17L, -17L, -17L, -17L, -17L, -17L, -16L, -17L, -17L, -17L, 
    -17L, -17L, -17L, -17L, -17L, -6L, 15L, 31L, 50L, 71L, 95L, 
    105L, 119L, 125L, 119L, 117L, 116L, 127L, 112L, 94L, 77L, 
    55L, 26L, 0L, -1L, -2L, 1L, -4L, -5L, -5L, -6L, -10L, -11L, 
    -11L, -11L, -10L, -9L, -10L, -8L, 4L, 9L, 15L, 25L, 41L, 
    57L, 71L, 84L, 100L, 117L, 133L, 134L, 137L, 134L, 134L, 
    135L, 139L, 135L, 131L, 120L, 125L, 132L, 131L, 127L, 140L, 
    143L, 144L, 143L, 143L, 142L, 139L, 138L, 131L, 124L, 104L, 
    80L, 61L, 45L, 23L, 9L, -4L, -4L, -8L, -13L, -13L, -13L, 
    -13L, -13L, -13L, -13L, -10L, -2L, 0L, 7L, 24L, 36L, 56L, 
    77L, 108L, 123L, 137L, 139L, 129L, 130L, 117L, 128L, 123L, 
    117L, 126L, 125L, 124L, 145L, 145L, 107L, 104L, 92L, 83L, 
    75L, 62L, 47L, 32L, 17L, 1L, -7L, -8L, -10L, -8L, -5L, -6L, 
    -3L, -1L, 10L, 20L, 28L, 32L, 33L, 34L, 52L, 74L, 98L, 114L, 
    116L, 116L, 121L, 122L, 123L, 120L, 117L, 114L, 107L, 109L, 
    87L, 69L, 37L, 12L, 8L, -14L, -14L, -14L, -11L, 2L, 17L, 
    29L, 25L, 12L, -9L, 0L, 12L, 28L, 54L, 84L, 96L, 105L, 114L, 
    114L, 111L, 112L, 111L, 116L, 120L, 120L, 127L), t = structure(1:294, .Label = c("0:00:00:00", 
    "0:00:00:14", "0:00:00:28", "0:00:00:42", "0:00:00:57", "0:00:00:71", 
    "0:00:00:85", "0:00:01:00", "0:00:01:14", "0:00:01:28", "0:00:01:42", 
    "0:00:01:57", "0:00:01:71", "0:00:01:85", "0:00:02:00", "0:00:02:14", 
    "0:00:02:28", "0:00:02:42", "0:00:02:57", "0:00:02:71", "0:00:02:85", 
    "0:00:03:00", "0:00:03:14", "0:00:03:28", "0:00:03:42", "0:00:03:57", 
    "0:00:03:71", "0:00:03:85", "0:00:04:00", "0:00:04:14", "0:00:04:28", 
    "0:00:04:42", "0:00:04:57", "0:00:04:71", "0:00:04:85", "0:00:05:00", 
    "0:00:05:14", "0:00:05:28", "0:00:05:42", "0:00:05:57", "0:00:05:71", 
    "0:00:05:85", "0:00:06:00", "0:00:06:14", "0:00:06:28", "0:00:06:42", 
    "0:00:06:57", "0:00:06:71", "0:00:06:85", "0:00:07:00", "0:00:07:14", 
    "0:00:07:28", "0:00:07:42", "0:00:07:57", "0:00:07:71", "0:00:07:85", 
    "0:00:08:00", "0:00:08:14", "0:00:08:28", "0:00:08:42", "0:00:08:57", 
    "0:00:08:71", "0:00:08:85", "0:00:09:00", "0:00:09:14", "0:00:09:28", 
    "0:00:09:42", "0:00:09:57", "0:00:09:71", "0:00:09:85", "0:00:10:00", 
    "0:00:10:14", "0:00:10:28", "0:00:10:42", "0:00:10:57", "0:00:10:71", 
    "0:00:10:85", "0:00:11:00", "0:00:11:14", "0:00:11:28", "0:00:11:42", 
    "0:00:11:57", "0:00:11:71", "0:00:11:85", "0:00:12:00", "0:00:12:14", 
    "0:00:12:28", "0:00:12:42", "0:00:12:57", "0:00:12:71", "0:00:12:85", 
    "0:00:13:00", "0:00:13:14", "0:00:13:28", "0:00:13:42", "0:00:13:57", 
    "0:00:13:71", "0:00:13:85", "0:00:14:00", "0:00:14:14", "0:00:14:28", 
    "0:00:14:42", "0:00:14:57", "0:00:14:71", "0:00:14:85", "0:00:15:00", 
    "0:00:15:14", "0:00:15:28", "0:00:15:42", "0:00:15:57", "0:00:15:71", 
    "0:00:15:85", "0:00:16:00", "0:00:16:14", "0:00:16:28", "0:00:16:42", 
    "0:00:16:57", "0:00:16:71", "0:00:16:85", "0:00:17:00", "0:00:17:14", 
    "0:00:17:28", "0:00:17:42", "0:00:17:57", "0:00:17:71", "0:00:17:85", 
    "0:00:18:00", "0:00:18:14", "0:00:18:28", "0:00:18:42", "0:00:18:57", 
    "0:00:18:71", "0:00:18:85", "0:00:19:00", "0:00:19:14", "0:00:19:28", 
    "0:00:19:42", "0:00:19:57", "0:00:19:71", "0:00:19:85", "0:00:20:00", 
    "0:00:20:14", "0:00:20:28", "0:00:20:42", "0:00:20:57", "0:00:20:71", 
    "0:00:20:85", "0:00:21:00", "0:00:21:14", "0:00:21:28", "0:00:21:42", 
    "0:00:21:57", "0:00:21:71", "0:00:21:85", "0:00:22:00", "0:00:22:14", 
    "0:00:22:28", "0:00:22:42", "0:00:22:57", "0:00:22:71", "0:00:22:85", 
    "0:00:23:00", "0:00:23:14", "0:00:23:28", "0:00:23:42", "0:00:23:57", 
    "0:00:23:71", "0:00:23:85", "0:00:24:00", "0:00:24:14", "0:00:24:28", 
    "0:00:24:42", "0:00:24:57", "0:00:24:71", "0:00:24:85", "0:00:25:00", 
    "0:00:25:14", "0:00:25:28", "0:00:25:42", "0:00:25:57", "0:00:25:71", 
    "0:00:25:85", "0:00:26:00", "0:00:26:14", "0:00:26:28", "0:00:26:42", 
    "0:00:26:57", "0:00:26:71", "0:00:26:85", "0:00:27:00", "0:00:27:14", 
    "0:00:27:28", "0:00:27:42", "0:00:27:57", "0:00:27:71", "0:00:27:85", 
    "0:00:28:00", "0:00:28:14", "0:00:28:28", "0:00:28:42", "0:00:28:57", 
    "0:00:28:71", "0:00:28:85", "0:00:29:00", "0:00:29:14", "0:00:29:28", 
    "0:00:29:42", "0:00:29:57", "0:00:29:71", "0:00:29:85", "0:00:30:00", 
    "0:00:30:14", "0:00:30:28", "0:00:30:42", "0:00:30:57", "0:00:30:71", 
    "0:00:30:85", "0:00:31:00", "0:00:31:14", "0:00:31:28", "0:00:31:42", 
    "0:00:31:57", "0:00:31:71", "0:00:31:85", "0:00:32:00", "0:00:32:14", 
    "0:00:32:28", "0:00:32:42", "0:00:32:57", "0:00:32:71", "0:00:32:85", 
    "0:00:33:00", "0:00:33:14", "0:00:33:28", "0:00:33:42", "0:00:33:57", 
    "0:00:33:71", "0:00:33:85", "0:00:34:00", "0:00:34:14", "0:00:34:28", 
    "0:00:34:42", "0:00:34:57", "0:00:34:71", "0:00:34:85", "0:00:35:00", 
    "0:00:35:14", "0:00:35:28", "0:00:35:42", "0:00:35:57", "0:00:35:71", 
    "0:00:35:85", "0:00:36:00", "0:00:36:14", "0:00:36:28", "0:00:36:42", 
    "0:00:36:57", "0:00:36:71", "0:00:36:85", "0:00:37:00", "0:00:37:14", 
    "0:00:37:28", "0:00:37:42", "0:00:37:57", "0:00:37:71", "0:00:37:85", 
    "0:00:38:00", "0:00:38:14", "0:00:38:28", "0:00:38:42", "0:00:38:57", 
    "0:00:38:71", "0:00:38:85", "0:00:39:00", "0:00:39:14", "0:00:39:28", 
    "0:00:39:42", "0:00:39:57", "0:00:39:71", "0:00:39:85", "0:00:40:00", 
    "0:00:40:14", "0:00:40:28", "0:00:40:42", "0:00:40:57", "0:00:40:71", 
    "0:00:40:85", "0:00:41:00", "0:00:41:14", "0:00:41:28", "0:00:41:42", 
    "0:00:41:57", "0:00:41:71", "0:00:41:85"), class = "factor")), .Names = c("Caja", 
"x", "y", "t"), row.names = c(NA, 294L), class = "data.frame")

对于我使用的情节

ggplot(data=df,aes(x,y)) + facet_wrap(~Caja) +
  stat_density2d(aes(fill=..level..,alpha=..level..),geom='polygon',colour='black') +
  scale_fill_continuous(low="green",high="red") +
  geom_point()+guides(alpha="none")

当我仅绘制caja4数据时

ggplot(data=caja4,aes(x,y)) + facet_wrap(~Caja) +
  stat_density2d(aes(fill=..level..,alpha=..level..),geom='polygon',colour='black') + 
  scale_fill_continuous(low="green",high="red") +
  geom_point()+guides(alpha="none")

这使得解释更容易，但如果你看一下线图，事情会有点不同

ggplot(caja4,aes(x,y))+geom_point()+geom_path(col="red")+theme_classic()

我想为小平面提供自由密度比例，但有一些东西更依赖于线图。

Answer 1

一般性评论

好的，我想在这里讨论一些事情：

您可以在方面调用中使用scales = 'free'，以便更好地将绘图空间与数据对齐。
您无法改变构面之间的密度比例。这是因为facet通常意味着使用 common 美学映射来表示不同的组。您可以改为制作单独的数字，并将它们粘在一起（例如cowplot包）。
多边形会在你的情节中被切断，让你的结果令人失望。我更愿意使用geom_tile来显示点的密度
正如MLavoie指出的那样，您需要使用geom_path代替geom_line。
那个色标并不是那么大。例如，它对色盲人的表现非常差。我建议使用viridis色标（见下文）。
如果x和y轴的单位相同且有意义（厘米或像素），您可以确保它们以相同的方式表示。您可以使用coord_equal()。

为一次试验制作一个漂亮的情节

下面是一些示例代码和一个用于组合的示例（我只有Caja 4的数据）：

library(viridis)

ggplot(caja4, aes(x,y)) +
  stat_density2d(geom = 'tile', aes(fill = ..density..), contour = FALSE) +
  geom_point() + 
  geom_path() +
  scale_fill_viridis() +
  coord_equal() +
  theme_minimal()

如果您确实想要使用轮廓多边形，则必须手动展开轴以避免剪切多边形（or you have close the polygons yourself）：

ggplot(caja4, aes(x,y)) +
  stat_density2d(geom = 'polygon', aes(fill = ..level..), contour = TRUE) +
  geom_point() + 
  geom_path() +
  scale_fill_viridis() +
  coord_equal() +
  theme_minimal() +
  xlim(-60, 40) +
  ylim(-50, 170)

结合不同的试验

我们可以使用数字面板代替facet（参见上面的第2点）。我发现cowplot包对此非常有用。下面是一些轻松创建复合图形的示例代码。（请注意，我在这里使用了log(..density..)来进行不同的颜色缩放。）

# Make data with three cajas (which only differ in x and y range here)
library(dplyr)

xy <- select(caja4, x, y)
cajas <- bind_rows('1' = xy, '2' = xy * 2, '3' = xy / 2, .id = 'Caja')

# Create ggplot objects for each caja
plots <- lapply(split(cajas, cajas$Caja),
                function(XX) {
                  plot = ggplot(XX, aes(x,y)) +
                    stat_density2d(geom = 'tile', aes(fill = log(..density..)), contour = FALSE) +
                    geom_point() + 
                    geom_path() +
                    scale_fill_viridis() +
                    coord_equal() +
                    theme_minimal()
                } )

plot_grid(plotlist = plots, nrow = 1, labels = 1:length(plots))

结果是：

2dDensity facet scale

1 个答案:

一般性评论

为一次试验制作一个漂亮的情节

结合不同的试验