Question

我在R中有这样一个数据帧：

set.seed(10)
sample <- data_frame(Date   = c('2000-05-01','2000-05-02','2000-05-03','2000-05-04',"2000-05-05",'2000-05-06'),
                 T1 = rnorm(6),
                 T2 = rnorm(6),
                 T3 = rnorm(6),
                 T1_a  = rnorm(6),
                 T1_b = rnorm(6), 
                 T1_c = rnorm(6), 
                 T2_a  = rnorm(6),
                 T2_b = rnorm(6), 
                 T2_c = rnorm(6), 
                 T3_a  = rnorm(6),
                 T3_b = rnorm(6), 
                 T3_c = rnorm(6))

我想使用MLmetrics包来计算均方根误差：

library(MLmetrics)
RMSE_T1_a = RMSE(sample$T1, sample$T1_a)
RMSE_T1_b = RMSE(sample$T1, sample$T1_b)
RMSE_T1_c = RMSE(sample$T1, sample$T1_c)

RMSE_T2_a = RMSE(sample$T2, sample$T2_a)
RMSE_T2_b = RMSE(sample$T2, sample$T2_b)
RMSE_T2_c = RMSE(sample$T2, sample$T2_c)

RMSE_T3_a = RMSE(sample$T3, sample$T3_a)
RMSE_T3_b = RMSE(sample$T3, sample$T3_b)
RMSE_T3_c = RMSE(sample$T3, sample$T3_c)

最后，我要将所有这些RMSE放在一个数据框中。

是否有更快的方式一次完成所有这些操作？

Answer 1

一种方法是像这样使用dplyr包：

library(dplyr)
rmsedata <- sample %>%
  summarise_at(vars(matches("T1_")), ~RMSE(T1, .x)) %>%
  bind_cols(sample %>%summarise_at(vars(matches("T2_")), ~RMSE(T2, .x))) %>%
  bind_cols(sample %>%summarise_at(vars(matches("T3_")), ~RMSE(T3, .x)))

      T1_a      T1_b    T1_c     T2_a      T2_b      T2_c      T3_a     T3_b     T3_c
1 1.391521 0.6828504 1.61983 1.195112 0.8101942 0.8953161 0.7983381 1.396028 1.171313

一个小建议：不要命名您的数据样本：有一个名为sample的R函数，同时命名您的数据样本也会引起混乱：）

计算R中特殊列的RMSE

1 个答案: