使用数据表来屏蔽/过滤 R 中的另一个数据表

Question

我有一个数据表，A 喜欢：

year          location    sigma_NN_1 sigma_NN_2 sigma_NN_3
2076 43.59375_-116.78125  1.4681173   1.664289   1.735974
2077 43.59375_-116.78125  1.3798515   1.550524   1.551269
2078 43.59375_-116.78125  0.7934367   1.064248   1.177981
2079 43.59375_-116.78125  1.8235574   1.991018   2.288402
2080 43.59375_-116.78125  2.5560329   2.578093   2.589334

我想用它来屏蔽另一个数据表，其中 sigmas 的值低于阈值，比如说 2。假设我的第二个数据表是 B

year            location       location_NN_1      location_NN_2      location_NN_3
2076 43.59375_-116.78125  41.15625_-90.65625 41.21875_-90.65625 41.15625_-90.65625
2077 43.59375_-116.78125  43.34375_-78.15625 43.34375_-78.21875 43.28125_-78.15625
2078 43.59375_-116.78125  41.34375_-90.78125 41.21875_-90.65625 41.53125_-73.96875
2079 43.59375_-116.78125 43.53125_-116.78125 41.34375_-90.78125 41.71875_-74.15625
2080 43.59375_-116.78125  41.34375_-90.78125 41.96875_-86.21875 41.21875_-90.65625

所以，我想要B[A<2] 之类的东西，但显然这不起作用，否则我不会在这里。

有什么建议吗？

预期输出：

输出

year            location       location_NN_1      location_NN_2      location_NN_3
2076 43.59375_-116.78125  41.15625_-90.65625 41.21875_-90.65625 41.15625_-90.65625
2077 43.59375_-116.78125  43.34375_-78.15625 43.34375_-78.21875 43.28125_-78.15625
2078 43.59375_-116.78125  41.34375_-90.78125 41.21875_-90.65625 41.53125_-73.96875
2079 43.59375_-116.78125 43.53125_-116.78125 41.34375_-90.78125                 NA
2080 43.59375_-116.78125                  NA                 NA                 NA

目标是找到数据表A中对应sigma小于2的位置。

Answer 1

我们可以使用基本 R 子集来识别B 的适当单元格，并将它们替换为NA。此方法要求A 和B 中的列顺序相同。

我们可以在dfa 上使用一个简单的条件语句来查找 sigma 值不小于 2 的单元格。由于我们不想将条件应用于年份和条件列，因此我们在应用之前将它们子集条件：

!(dfa[-c(1,2)] < 2)
     sigma_NN_1 sigma_NN_2 sigma_NN_3
[1,]      FALSE      FALSE      FALSE
[2,]      FALSE      FALSE      FALSE
[3,]      FALSE      FALSE      FALSE
[4,]      FALSE      FALSE       TRUE
[5,]       TRUE       TRUE       TRUE

这会返回一个逻辑矩阵，我们可以使用它来对B 进行子集化以替换值。这里发生的是我们将B 子集两次：首先我们忽略年份和位置列以仅获取位置列，然后我们使用前面的条件选择匹配sigma 值不小于2 的行并插入@987654329 @进去：

dfb[-c(1,2)][!(dfa[-c(1,2)] < 2)] <- NA
dfb

  year            location       location_NN_1      location_NN_2      location_NN_3
1 2076 43.59375_-116.78125  41.15625_-90.65625 41.21875_-90.65625 41.15625_-90.65625
2 2077 43.59375_-116.78125  43.34375_-78.15625 43.34375_-78.21875 43.28125_-78.15625
3 2078 43.59375_-116.78125  41.34375_-90.78125 41.21875_-90.65625 41.53125_-73.96875
4 2079 43.59375_-116.78125 43.53125_-116.78125 41.34375_-90.78125               <NA>
5 2080 43.59375_-116.78125                <NA>               <NA>               <NA>

Answer 2

简单的基础 R 解决方案：

B[-(1:2)][A[-(1:2)]>=2] <- NA

选择除第一和第二B[-(1:2)] 之外的所有列。

然后使用矢量化逻辑表达式A[-(1:2)]>=2 将正确的元素设置为NA。

结果：

  year            location       location_NN_1      location_NN_2      location_NN_3
1 2076 43.59375_-116.78125  41.15625_-90.65625 41.21875_-90.65625 41.15625_-90.65625
2 2077 43.59375_-116.78125  43.34375_-78.15625 43.34375_-78.21875 43.28125_-78.15625
3 2078 43.59375_-116.78125  41.34375_-90.78125 41.21875_-90.65625 41.53125_-73.96875
4 2079 43.59375_-116.78125 43.53125_-116.78125 41.34375_-90.78125               <NA>
5 2080 43.59375_-116.78125                <NA>               <NA>               <NA>

Answer 3

假设这些data.table对象，并假设'A'中'sigma'列的行数都应该小于阈值2。

library(data.table)
nm1 <- grep("sigma", names(A), value = TRUE)
i1 <- setDT(A)[, Reduce(`&`, lapply(.SD, `<`, 2)), .SDcols = nm1]
setDT(B)[i1] 

更新

基于预期输出

nm2 <- grep("sigma", names(A))
B[, (nm2) := Map(function(x, y) replace(x, y >= 2, NA_character_),
        .SD, A[, nm2, with = FALSE]), .SDcols = nm2][]
# year            location       location_NN_1      location_NN_2      location_NN_3
#1: 2076 43.59375_-116.78125  41.15625_-90.65625 41.21875_-90.65625 41.15625_-90.65625
#2: 2077 43.59375_-116.78125  43.34375_-78.15625 43.34375_-78.21875 43.28125_-78.15625
#3: 2078 43.59375_-116.78125  41.34375_-90.78125 41.21875_-90.65625 41.53125_-73.96875
#4: 2079 43.59375_-116.78125 43.53125_-116.78125 41.34375_-90.78125               <NA>
#5: 2080 43.59375_-116.78125                <NA>               <NA>               <NA>

数据

A <- structure(list(year = 2076:2080, location = c("43.59375_-116.78125", 
"43.59375_-116.78125", "43.59375_-116.78125", "43.59375_-116.78125", 
"43.59375_-116.78125"), sigma_NN_1 = c(1.4681173, 1.3798515, 
0.7934367, 1.8235574, 2.5560329), sigma_NN_2 = c(1.664289, 1.550524, 
1.064248, 1.991018, 2.578093), sigma_NN_3 = c(1.735974, 1.551269, 
1.177981, 2.288402, 2.589334)), class = "data.frame", row.names = c(NA, 
-5L))

B <- structure(list(year = 2076:2080, location = c("43.59375_-116.78125", 
"43.59375_-116.78125", "43.59375_-116.78125", "43.59375_-116.78125", 
"43.59375_-116.78125"), location_NN_1 = c("41.15625_-90.65625", 
"43.34375_-78.15625", "41.34375_-90.78125", "43.53125_-116.78125", 
"41.34375_-90.78125"), location_NN_2 = c("41.21875_-90.65625", 
"43.34375_-78.21875", "41.21875_-90.65625", "41.34375_-90.78125", 
"41.96875_-86.21875"), location_NN_3 = c("41.15625_-90.65625", 
"43.28125_-78.15625", "41.53125_-73.96875", "41.71875_-74.15625", 
"41.21875_-90.65625")), class = "data.frame", row.names = c(NA, 
-5L))