欠損値、欠損データ

Rでの欠損データの扱い方

Nicholas Tierney

Statistician

欠損データの別の見方

name	time	value
robin	morning	358
robin	afternoon	534
robin	evening	100
sam	morning	139
sam	afternoon	177
blair	morning	963
blair	afternoon	962
blair	evening	929

name	afternoon	evening	morning
blair	962	929	963
robin	534	100	358
sam	177	NA	139

明示的・暗黙的な欠損値

明示的：NA として表示される
暗黙的：データに表示されないが、存在が示唆される

暗黙的な欠損値を明示化する

tetris %>%
  tidyr::complete(name, time)

# A tibble: 9 x 3
  name  time      value
  <fct> <fct>     <dbl>
1 blair afternoon   962
2 blair evening     929
3 blair morning     963
4 robin afternoon   534
5 robin evening     100
6 robin morning     358
7 sam   afternoon   177
8 sam   evening      NA
9 sam   morning     139

明示的な欠損値の処理

name	time	value
robin	morning	936
NA	afternoon	635
NA	evening	438
sam	morning	208
NA	afternoon	92
NA	evening	79
blair	morning	969
NA	afternoon	918
NA	evening	954

name	time	value
robin	morning	936
robin	afternoon	635
robin	evening	438
sam	morning	208
sam	afternoon	92
sam	evening	79
blair	morning	969
blair	afternoon	918
blair	evening	954

明示的な欠損値の処理

name	time	value
robin	morning	936
NA	afternoon	635
NA	evening	438
sam	morning	208
NA	afternoon	92
NA	evening	79
blair	morning	969
NA	afternoon	918
NA	evening	954

tetris %>%
  tidyr::fill(name)

# A tibble: 9 x 3
  name  time      value
  <chr> <chr>     <dbl>
1 robin morning     936
2 robin afternoon   635
3 robin evening     438
4 sam   morning     208
5 sam   afternoon    92
6 sam   evening      79
7 blair morning     969
8 blair afternoon   918
9 blair evening     954

注意事項

tetris %>%
  tidyr::fill(name)

# A tibble: 9 x 3
  name  time      value
  <chr> <chr>     <dbl>
1 robin morning     936
2 robin afternoon   635
3 robin evening     438
4 sam   morning     208
5 sam   afternoon    92
6 sam   evening      79
7 blair morning     969
8 blair afternoon   918
9 blair evening     954

では、練習しましょう！

Rでの欠損データの扱い方