使用 R 數據整理將一列拆分為兩列

Question

我非常感謝您在使用 R 進行數據整理方面的幫助。 我有一個數據，我想在適用的情況下將一列（變量）拆分為兩列（以其他變量為條件）。 例如，根據下面的示例，數據表示在不同閱讀時間（塊）中出現的某些單詞（項目）的反應時間度量（RT1 和 RT2）。 我想看看塊3、4和5中的RT1和RT2值是否與塊1中相同項目的RT1和RT2值相關。在塊1中出現並在后續塊中重新出現的目標項目被編碼為“條件”列中的“EI”，而編碼為“E”或“I”的項目僅出現一次。

dput(d1)
structure(list(RECORDING_SESSION_LABEL = c(26, 26, 26, 26, 26, 
26, 26, 26), RT1 = c(5171, 3857, 3447, 314, 460, 731, 957, 1253
), RT2 = c(357, 328, 122, 39, 86, 132, 173, 215), item = c("foreign", 
"detailed", "large", "foreign", "foreign", "large", "large", 
"disputable"), block = c(1, 1, 1, 3, 4, 3, 4, 3), condition = c("EI", 
"E", "EI", "EI", "EI", "EI", "EI", "I")), row.names = c(NA, -8L
), class = c("tbl_df", "tbl", "data.frame"))

數據樣本如下所示：

> d1
# A tibble: 8 x 6
  RECORDING_SESSION_LABEL   RT1   RT2 item       block condition
                    <dbl> <dbl> <dbl> <chr>      <dbl> <chr>    
1                      26  5171   357 foreign        1 EI       
2                      26  3857   328 detailed       1 E        
3                      26  3447   122 large          1 EI       
4                      26   314    39 foreign        3 EI       
5                      26   460    86 foreign        4 EI       
6                      26   731   132 large          3 EI       
7                      26   957   173 large          4 EI       
8                      26  1253   215 disputable     3 I   

為了以 R 可以理解的格式呈現，我想要實現的目標數據框將類似於下面的（應添加突出顯示的列）。 這些列中的空白行表示不重復出現的項目（條件未編碼為“EI”）； 因此，它們無關緊要，應編碼為“NA”。

dput(d2)
structure(list(RECORDING_SESSION_LABEL = c(26, 26, 26, 26, 26, 
26, 26, 26), `RT 1` = c(5171, 3857, 3447, 314, 460, 731, 957, 
1253), RT2 = c(357, 328, 122, 39, 86, 132, 173, 215), item = c("foreign", 
"detailed", "large", "foreign", "foreign", "large", "large", 
"disputable"), block = c(1, 1, 1, 3, 4, 3, 4, 3), condition = c("EI", 
"E", "EI", "EI", "EI", "EI", "EI", "I"), `RT 1_at_block1` = c(NA, 
NA, NA, 5171, 5171, 3447, 3447, NA), RT2_at_block1 = c(NA, NA, 
NA, 357, 357, 122, 122, NA)), row.names = c(NA, -8L), class = c("tbl_df", 
"tbl", "data.frame"))

目標數據格式的示例如下所示：

> d2
# A tibble: 8 x 8
  RECORDING_SESSI~ `RT 1`   RT2 item  block condition `RT 1_at_block1`
             <dbl>  <dbl> <dbl> <chr> <dbl> <chr>                <dbl>
1               26   5171   357 fore~     1 EI                      NA
2               26   3857   328 deta~     1 E                       NA
3               26   3447   122 large     1 EI                      NA
4               26    314    39 fore~     3 EI                    5171
5               26    460    86 fore~     4 EI                    5171
6               26    731   132 large     3 EI                    3447
7               26    957   173 large     4 EI                    3447
8               26   1253   215 disp~     3 I                       NA
# ... with 1 more variable: RT2_at_block1 <dbl>

> head(d2)
# A tibble: 6 x 8
  RECORDING_SESSION_LABEL `RT 1`   RT2 item     block condition `RT 1_at_block1` RT2_at_block1
                    <dbl>  <dbl> <dbl> <chr>    <dbl> <chr>                <dbl>         <dbl>
1                      26   5171   357 foreign      1 EI                      NA            NA
2                      26   3857   328 detailed     1 E                       NA            NA
3                      26   3447   122 large        1 EI                      NA            NA
4                      26    314    39 foreign      3 EI                    5171           357
5                      26    460    86 foreign      4 EI                    5171           357
6                      26    731   132 large        3 EI                    3447           122

在此先感謝您的幫助。

Answer 1

使用dplyr的可能解決方案：

  d1 <- structure(list(RECORDING_SESSION_LABEL = c(26, 26, 26, 26, 26, 26, 26, 26),
                       RT1 = c(5171, 3857, 3447, 314, 460, 731, 957, 1253),
                       RT2 = c(357, 328, 122, 39, 86, 132, 173, 215),
                       item = c("foreign", "detailed", "large", "foreign", "foreign", "large", "large", "disputable"),
                       block = c(1, 1, 1, 3, 4, 3, 4, 3), condition = c("EI", "E", "EI", "EI", "EI", "EI", "EI", "I")),
                  row.names = c(NA, -8L), class = c("tbl_df", "tbl", "data.frame"))



  library(dplyr)

  d2 <- d1 %>% 
    left_join(d1 %>% filter(block == 1) %>% select(RECORDING_SESSION_LABEL, item, RT1_at_block1 = RT1)) %>% 
    left_join(d1 %>% filter(block == 1) %>% select(RECORDING_SESSION_LABEL, item, RT2_at_block1 = RT2))

之后， d2看起來像這樣：

    RECORDING_SESSION_LABEL   RT1   RT2 item     block condition RT1_at_block1 RT2_at_block1
                      <dbl> <dbl> <dbl> <chr>    <dbl> <chr>             <dbl>         <dbl>
  1                      26  5171   357 foreign      1 EI                 5171           357
  2                      26  3857   328 detailed     1 E                  3857           328
  3                      26  3447   122 large        1 EI                 3447           122
  4                      26   314    39 foreign      3 EI                 5171           357
  5                      26   460    86 foreign      4 EI                 5171           357
  6                      26   731   132 large        3 EI                 3447           122

編輯：如果要將塊 1 的值設置為NA添加mutate ：

d2 <- d1 %>% 
  left_join(d1 %>% filter(block == 1) %>% select(RECORDING_SESSION_LABEL, item, RT1_at_block1 = RT1)) %>% 
  left_join(d1 %>% filter(block == 1) %>% select(RECORDING_SESSION_LABEL, item, RT2_at_block1 = RT2)) %>% 
  mutate(RT1_at_block1 = ifelse(block == 1, NA, RT1_at_block1),
         RT2_at_block1 = ifelse(block == 1, NA, RT2_at_block1))