如何按组在R中进行多项算术运算
How to do multiple arithmetic operations in R by group
有几个数据集。第一个
lid=structure(list(x1 = 619490L, x2 = 10L, x3 = 0L, x4 = 6089230L,
x5 = 0L, x6 = -10L), class = "data.frame", row.names = c(NA,
-1L))
第二个数据集
lidar=structure(list(A = c(638238.76, 638238.76, 638239.29, 638235.39,
638233.86, 638233.86, 638235.55, 638231.97, 638231.91, 638228.41,
638238.76, 638238.76, 63239.29, 638235.39, 638233.86, 638233.86,
638235.55, 638231.97, 638231.91, 638228.41), B = c(6078001.09,
6078001.09, 6078001.15, 6078001.15, 6078001.07, 6078001.07, 6078001.02,
6078001.08, 6078001.09, 6078001.01, 6078001.09, 6078001.09, 6078001.15,
6078001.15, 6078001.07, 6078001.07, 6078001.02, 6078001.08, 6078001.09,
6078001.01), C = c(186.64, 186.59, 199.28, 189.37, 186.67, 186.67,
198.04, 200.03, 199.73, 192.14, 186.64, 186.59, 199.28, 189.37,
196.67, 186.67, 198.04, 200.03, 199.73, 100.14), gpstime = c(319805734.664265,
319805734.664265, 319805734.67875, 319805734.678768, 319805734.678777,
319805734.678777, 319805734.687338, 319805734.701928, 319805734.701928,
319805734.701945, 319805734.664265, 319805734.664265, 319805734.67875,
319805734.678768, 319805734.678777, 319805734.678777, 319805734.687338,
319805734.701928, 319805734.701928, 319805734.701945), Intensity = c(13L,
99L, 5L, 2L, 20L, 189L, 2L, 11L, 90L, 1L, 13L, 99L, 5L, 2L, 20L,
189L, 2L, 11L, 90L, 1L), ReturnNumber = c(2L, 1L, 1L, 2L, 1L,
1L, 2L, 1L, 1L, 3L, 2L, 1L, 1L, 2L, 1L, 1L, 2L, 1L, 1L, 3L),
NumberOfReturns = c(2L, 1L, 3L, 2L, 1L, 1L, 3L, 1L, 1L, 4L,
2L, 1L, 3L, 2L, 1L, 1L, 3L, 1L, 1L, 4L), ScanDirectionFlag = c(1L,
1L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 1L, 1L, 0L, 0L, 0L, 0L,
1L, 0L, 0L, 0L), EdgeOfFlightline = c(0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L
), Classification = c(1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), group = c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L)), class = "data.frame", row.names = c(NA,
-20L))
lid 静态数据集，它总是相同的（总是相同的值）。
我必须执行这些算术运算
` lidar$row <- round((lidar$A-lid$x1)/lid$x3, 0)
lidar$col <- (lidar$B-lid$x4)/lid$x6
lidar$cdif <- max(lidar$C)-min(lidar$C)
但每个激光雷达 $groups 分别。
如何使用 dplyr 做得更好？
感谢您的帮助。
library(dplyr)
lidar %>%
group_by(group) %>%
mutate(
row = (A-lid$x1)/lid$x3,
col = (B-lid$x4)/lid$x6,
cdif = max(C)-min(C)
) %>%
ungroup()
# # A tibble: 20 x 14
# A B C gpstime Intensity ReturnNumber NumberOfReturns ScanDirectionFlag EdgeOfFlightline Classification group row col cdif
# <dbl> <dbl> <dbl> <dbl> <int> <int> <int> <int> <int> <int> <int> <dbl> <dbl> <dbl>
# 1 638239. 6078001. 187. 319805735. 13 2 2 1 0 1 1 Inf 1123. 13.4
# 2 638239. 6078001. 187. 319805735. 99 1 1 1 0 2 1 Inf 1123. 13.4
# 3 638239. 6078001. 199. 319805735. 5 1 3 0 0 1 1 Inf 1123. 13.4
# 4 638235. 6078001. 189. 319805735. 2 2 2 0 0 1 1 Inf 1123. 13.4
# 5 638234. 6078001. 187. 319805735. 20 1 1 0 0 1 1 Inf 1123. 13.4
# 6 638234. 6078001. 187. 319805735. 189 1 1 0 0 1 1 Inf 1123. 13.4
# 7 638236. 6078001. 198. 319805735. 2 2 3 1 0 1 1 Inf 1123. 13.4
# 8 638232. 6078001. 200. 319805735. 11 1 1 0 0 1 1 Inf 1123. 13.4
# 9 638232. 6078001. 200. 319805735. 90 1 1 0 0 1 1 Inf 1123. 13.4
# 10 638228. 6078001. 192. 319805735. 1 3 4 0 0 1 1 Inf 1123. 13.4
# 11 638239. 6078001. 187. 319805735. 13 2 2 1 0 1 2 Inf 1123. 99.9
# 12 638239. 6078001. 187. 319805735. 99 1 1 1 0 2 2 Inf 1123. 99.9
# 13 63239. 6078001. 199. 319805735. 5 1 3 0 0 1 2 -Inf 1123. 99.9
# 14 638235. 6078001. 189. 319805735. 2 2 2 0 0 1 2 Inf 1123. 99.9
# 15 638234. 6078001. 197. 319805735. 20 1 1 0 0 1 2 Inf 1123. 99.9
# 16 638234. 6078001. 187. 319805735. 189 1 1 0 0 1 2 Inf 1123. 99.9
# 17 638236. 6078001. 198. 319805735. 2 2 3 1 0 1 2 Inf 1123. 99.9
# 18 638232. 6078001. 200. 319805735. 11 1 1 0 0 1 2 Inf 1123. 99.9
# 19 638232. 6078001. 200. 319805735. 90 1 1 0 0 1 2 Inf 1123. 99.9
# 20 638228. 6078001. 100. 319805735. 1 3 4 0 0 1 2 Inf 1123. 99.9
row 总是 Inf 因为 lid$x3 是 0。唯一必须分组的部分是 xdif，因为它是唯一进行分组聚合的部分，其余部分可以不分组地完成。
lidar %>%
mutate(
row = (A-lid$x1)/lid$x3,
col = (B-lid$x4)/lid$x6
) %>%
group_by(group) %>%
mutate(cdif = max(C)-min(C)) %>%
ungroup()
为什么要这样做？对于更大的数据集或有很多组，一次处理整个向量而不是 per-group 会更有效（可能明显更快）。实际计算应该return相同的结果。
有几个数据集。第一个
lid=structure(list(x1 = 619490L, x2 = 10L, x3 = 0L, x4 = 6089230L,
x5 = 0L, x6 = -10L), class = "data.frame", row.names = c(NA,
-1L))
第二个数据集
lidar=structure(list(A = c(638238.76, 638238.76, 638239.29, 638235.39,
638233.86, 638233.86, 638235.55, 638231.97, 638231.91, 638228.41,
638238.76, 638238.76, 63239.29, 638235.39, 638233.86, 638233.86,
638235.55, 638231.97, 638231.91, 638228.41), B = c(6078001.09,
6078001.09, 6078001.15, 6078001.15, 6078001.07, 6078001.07, 6078001.02,
6078001.08, 6078001.09, 6078001.01, 6078001.09, 6078001.09, 6078001.15,
6078001.15, 6078001.07, 6078001.07, 6078001.02, 6078001.08, 6078001.09,
6078001.01), C = c(186.64, 186.59, 199.28, 189.37, 186.67, 186.67,
198.04, 200.03, 199.73, 192.14, 186.64, 186.59, 199.28, 189.37,
196.67, 186.67, 198.04, 200.03, 199.73, 100.14), gpstime = c(319805734.664265,
319805734.664265, 319805734.67875, 319805734.678768, 319805734.678777,
319805734.678777, 319805734.687338, 319805734.701928, 319805734.701928,
319805734.701945, 319805734.664265, 319805734.664265, 319805734.67875,
319805734.678768, 319805734.678777, 319805734.678777, 319805734.687338,
319805734.701928, 319805734.701928, 319805734.701945), Intensity = c(13L,
99L, 5L, 2L, 20L, 189L, 2L, 11L, 90L, 1L, 13L, 99L, 5L, 2L, 20L,
189L, 2L, 11L, 90L, 1L), ReturnNumber = c(2L, 1L, 1L, 2L, 1L,
1L, 2L, 1L, 1L, 3L, 2L, 1L, 1L, 2L, 1L, 1L, 2L, 1L, 1L, 3L),
NumberOfReturns = c(2L, 1L, 3L, 2L, 1L, 1L, 3L, 1L, 1L, 4L,
2L, 1L, 3L, 2L, 1L, 1L, 3L, 1L, 1L, 4L), ScanDirectionFlag = c(1L,
1L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 1L, 1L, 0L, 0L, 0L, 0L,
1L, 0L, 0L, 0L), EdgeOfFlightline = c(0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L
), Classification = c(1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), group = c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L)), class = "data.frame", row.names = c(NA,
-20L))
lid 静态数据集，它总是相同的（总是相同的值）。
我必须执行这些算术运算
` lidar$row <- round((lidar$A-lid$x1)/lid$x3, 0)
lidar$col <- (lidar$B-lid$x4)/lid$x6
lidar$cdif <- max(lidar$C)-min(lidar$C)
但每个激光雷达 $groups 分别。
如何使用 dplyr 做得更好？
感谢您的帮助。
library(dplyr)
lidar %>%
group_by(group) %>%
mutate(
row = (A-lid$x1)/lid$x3,
col = (B-lid$x4)/lid$x6,
cdif = max(C)-min(C)
) %>%
ungroup()
# # A tibble: 20 x 14
# A B C gpstime Intensity ReturnNumber NumberOfReturns ScanDirectionFlag EdgeOfFlightline Classification group row col cdif
# <dbl> <dbl> <dbl> <dbl> <int> <int> <int> <int> <int> <int> <int> <dbl> <dbl> <dbl>
# 1 638239. 6078001. 187. 319805735. 13 2 2 1 0 1 1 Inf 1123. 13.4
# 2 638239. 6078001. 187. 319805735. 99 1 1 1 0 2 1 Inf 1123. 13.4
# 3 638239. 6078001. 199. 319805735. 5 1 3 0 0 1 1 Inf 1123. 13.4
# 4 638235. 6078001. 189. 319805735. 2 2 2 0 0 1 1 Inf 1123. 13.4
# 5 638234. 6078001. 187. 319805735. 20 1 1 0 0 1 1 Inf 1123. 13.4
# 6 638234. 6078001. 187. 319805735. 189 1 1 0 0 1 1 Inf 1123. 13.4
# 7 638236. 6078001. 198. 319805735. 2 2 3 1 0 1 1 Inf 1123. 13.4
# 8 638232. 6078001. 200. 319805735. 11 1 1 0 0 1 1 Inf 1123. 13.4
# 9 638232. 6078001. 200. 319805735. 90 1 1 0 0 1 1 Inf 1123. 13.4
# 10 638228. 6078001. 192. 319805735. 1 3 4 0 0 1 1 Inf 1123. 13.4
# 11 638239. 6078001. 187. 319805735. 13 2 2 1 0 1 2 Inf 1123. 99.9
# 12 638239. 6078001. 187. 319805735. 99 1 1 1 0 2 2 Inf 1123. 99.9
# 13 63239. 6078001. 199. 319805735. 5 1 3 0 0 1 2 -Inf 1123. 99.9
# 14 638235. 6078001. 189. 319805735. 2 2 2 0 0 1 2 Inf 1123. 99.9
# 15 638234. 6078001. 197. 319805735. 20 1 1 0 0 1 2 Inf 1123. 99.9
# 16 638234. 6078001. 187. 319805735. 189 1 1 0 0 1 2 Inf 1123. 99.9
# 17 638236. 6078001. 198. 319805735. 2 2 3 1 0 1 2 Inf 1123. 99.9
# 18 638232. 6078001. 200. 319805735. 11 1 1 0 0 1 2 Inf 1123. 99.9
# 19 638232. 6078001. 200. 319805735. 90 1 1 0 0 1 2 Inf 1123. 99.9
# 20 638228. 6078001. 100. 319805735. 1 3 4 0 0 1 2 Inf 1123. 99.9
row 总是 Inf 因为 lid$x3 是 0。唯一必须分组的部分是 xdif，因为它是唯一进行分组聚合的部分，其余部分可以不分组地完成。
lidar %>%
mutate(
row = (A-lid$x1)/lid$x3,
col = (B-lid$x4)/lid$x6
) %>%
group_by(group) %>%
mutate(cdif = max(C)-min(C)) %>%
ungroup()
为什么要这样做？对于更大的数据集或有很多组，一次处理整个向量而不是 per-group 会更有效（可能明显更快）。实际计算应该return相同的结果。