根据特定值按组向面板数据添加计数器

Question

我有一个面板数据集，需要在特定年份之前和之后添加一个计数器。即，我有以下数据：

my.panel <- data.frame(id =c(1,1,1,1,1,1,1,1,1, 2,2,2,2,2,2,2,2,2), year = c(1910, 1911, 1912, 1913, 1914, 1915, 1916, 1917, 1918, 1922, 1923, 1924, 1925, 1926, 1927, 1928, 1929, 1930)
                       , indicator = c(0,0,0,1,0,0,0,0,0, 0,0,0,0,0,0,1,0,0)) 
my.panel

   id year indicator
1   1 1910         0
2   1 1911         0
3   1 1912         0
4   1 1913         1
5   1 1914         0
6   1 1915         0
7   1 1916         0
8   1 1917         0
9   1 1918         0
10  2 1922         0
11  2 1923         0
12  2 1924         0
13  2 1925         0
14  2 1926         0
15  2 1927         0
16  2 1928         1
17  2 1929         0
18  2 1930         0

并且需要以下内容：

   id year indicator counter
1   1 1910         0      -3
2   1 1911         0      -2
3   1 1912         0      -1
4   1 1913         1       0
5   1 1914         0       1
6   1 1915         0       2
7   1 1916         0       3
8   1 1917         0       4
9   1 1918         0       5
10  2 1922         0      -6
11  2 1923         0      -5
12  2 1924         0      -4
13  2 1925         0      -3
14  2 1926         0      -2
15  2 1927         0      -1
16  2 1928         1       0
17  2 1929         0       1
18  2 1930         0       2

我敢打赌那里有一些简单的dplyr解决方案。

Answer 1

使用dplyr：

my.panel %>%
  group_by(id) %>%
  mutate(counter=year-year[indicator==1])

此产量

# A tibble: 18 x 4
# Groups:   id [2]
      id  year indicator counter
   <dbl> <dbl>     <dbl>   <dbl>
 1     1  1910         0      -3
 2     1  1911         0      -2
 3     1  1912         0      -1
 4     1  1913         1       0
 5     1  1914         0       1
 6     1  1915         0       2
 7     1  1916         0       3
 8     1  1917         0       4
 9     1  1918         0       5
10     2  1922         0      -6
11     2  1923         0      -5
12     2  1924         0      -4
13     2  1925         0      -3
14     2  1926         0      -2
15     2  1927         0      -1
16     2  1928         1       0
17     2  1929         0       1
18     2  1930         0       2