用R中缺失的数据填充数据

Question

我有一个代码，计算10x10的每个象限有多少点，并在每个象限中显示结果。

我有这个数据

      qx    qy     n
   <dbl> <dbl> <int>
 1    10    80     1
 2    10    90     1
 3    20    20     1
 4    20    30     1
 5    20    40     1
 6    20    60     3
 7    30    10     2
 8    30    20     1
 9    30    30     1
10    30    40     2
# ... with 38 more rows

我就是这样创建图形的

ggplot() +
    scale_x_continuous(breaks = seq(0, 100, by = 10)) +
    scale_y_continuous(breaks = seq(0, 100, by = 10)) +
    geom_text(data=df, mapping=aes(x=qx-5, y=qy-5, label=n, size=1))

如何将丢失的象限添加到数据帧中？

失踪象限：

 10    10     0
 20    10     0
 70    10     0
 ...etc

Answer 1

用所有网格位置创建一个新的空数据框架，然后用已知的值填充可能会更容易。

详情见评论：

library(dplyr)
#create default matrix with all grid locations
qx <- rep(seq(10, 90, 10), each=9)
qy <- rep(seq(10, 90, 10), 9)
empty <- data.frame(qx, qy)

data<- read.table(header=TRUE, text=" qx    qy     n
 1    10    80     1
 2    10    90     1
 3    20    20     1
 4    20    30     1
 5    20    40     1
 6    20    60     3
 7    30    10     2
 8    30    20     1
 9    30    30     1
10    30    40     2")

#merge the known data with the master matrix
answer <-left_join(empty, data, by=c("qx"="qx", "qy"="qy"))
#replace NA with 0
answer$n[is.na(answer$n)] <-0

Answer 2

按照与@Dave2e相同的逻辑，我首先使用tidyverse为网格位置创建包含所有组合的数据，然后加入数据，将NA值替换为0，然后绘制。由于我只使用数据的一个子集，那么值将不与问题中的网格相同。

library(tidyverse)

crossing(qx = seq(10, 100, 10), qy = seq(10, 100, 10)) %>% 
  left_join(., df, by = c("qx", "qy")) %>% 
  mutate(n = replace_na(n, 0)) %>% 
  ggplot(.) +
  scale_x_continuous(breaks = seq(0, 100, by = 10)) +
  scale_y_continuous(breaks = seq(0, 100, by = 10)) +
  geom_text(mapping = aes(x = qx - 5, y = qy - 5, label = n, size = 1))

输出

​

​

数据

df <- structure(list(qx = c(10L, 10L, 20L, 20L, 20L, 20L, 30L, 30L, 
30L, 30L), qy = c(80L, 90L, 20L, 30L, 40L, 60L, 10L, 20L, 30L, 
40L), n = c(1L, 1L, 1L, 1L, 1L, 3L, 2L, 1L, 1L, 2L)), class = "data.frame", row.names = c("1", 
"2", "3", "4", "5", "6", "7", "8", "9", "10"))