R data.table：通过在另一列中找到的值重新设定面板中的每个组

Question

假设我的数据集如下所示：

data.table(groups = rep(c("A","B"), each=5), time = rep(2011:2015, 2), value = runif(10,95,105))

    groups time     value
 1:      A 2011  97.66647
 2:      A 2012 103.00917
 3:      A 2013 103.02462
 4:      A 2014 103.25105
 5:      A 2015 101.94881
 6:      B 2011  97.69248
 7:      B 2012  95.63296
 8:      B 2013  98.40150
 9:      B 2014 104.14164
10:      B 2015 103.61722

我想通过与所选年份（例如 2013 年）相对应的值来重新设定每个组内的值。 因此，对于 AI 组，希望将每个值除以 103.02462（2013 年的值），对于 B 组除以 98.40150，等等......

我能想到的所有解决方案都非常复杂，如果有人能分享他们的想法就好了

Answer 1

按'groups'分组后，得到对应于2013年'time'值的'value'，并用它来划分'value'列

library(data.table)
dt1[, value := value/value[time == 2013], by = groups]
dt1
#    groups time     value
# 1:      A 2011 0.9479916
# 2:      A 2012 0.9998500
# 3:      A 2013 1.0000000
# 4:      A 2014 1.0021978
# 5:      A 2015 0.9895577
# 6:      B 2011 0.9927946
# 7:      B 2012 0.9718649
# 8:      B 2013 1.0000000
# 9:      B 2014 1.0583339
#10:      B 2015 1.0530045

或者match

 dt1[, value := value/value[match(2013, time)], by = groups]

数据

dt1 <- structure(list(groups = c("A", "A", "A", "A", "A", "B", "B", 
"B", "B", "B"), time = c(2011L, 2012L, 2013L, 2014L, 2015L, 2011L, 
2012L, 2013L, 2014L, 2015L), value = c(97.66647, 103.00917, 103.02462, 
103.25105, 101.94881, 97.69248, 95.63296, 98.4015, 104.14164, 
103.61722)), class = c("data.table", "data.frame"), row.names = c(NA, 
-10L))