CODEWITHSUNDEEP
rdataframedplyr
Balachandar Kaliappan
Balachandar Kaliappan

Reputation: 153

Removing NA's using filter function on few columns of the data frame

I have a large data frame that has NA's at different point. I need to remove few rows that has more NA values.

I applied filter using is.na() conditions to remove them. However, they are not yielding fruitful results.

S.No    MediaName KeyPress     KPIndex  Type            Secs    X       Y
001     Dat       NA           1        Fixation        18      117     89
002     New       NA           NA       Saccade         33      NA      NA
003     Dat       NA           2        Fixation        23      117     NA

my code

df <- df%>%filter(df, !is.na(KeyPress) & !is.na(KPIndex) & !is.na(X) & !is.na(Y))

I would want delete based on the conditions using dplyr. I have more rows similar to this in a large dataframe. What is wrong with my code?

Upvotes: 2

Views: 996

Answers (3)

GenesRus
GenesRus

Reputation: 1057

I'll put my bid in for the simplest. Dplyr's drop_na will drop all rows that contain an NA. 

data %>% drop_na()

You can specify which columns you want it to look at (or not with "-"), if that is relevant to your case.

Upvotes: 0

akrun
akrun

Reputation: 886938

If there are more than one column, use filter_at

library(dplyr)     
df %>%
   filter_at(vars(KeyPress, KPIndex, X, Y), any_vars(!is.na(.)))

Or with rowSums from base R

nm1 <- c("KeyPress", "KPIndex", "X", "Y")
df[rowSums(!is.na(df[nm1]))!= 0,]

data

df <- structure(list(S.No = 1:3, MediaName = c("Dat", "New", "Dat"), 
    KeyPress = c(NA, NA, NA), KPIndex = c(1L, NA, 2L), Type = c("Fixation", 
    "Saccade", "Fixation"), Secs = c(18L, 33L, 23L), X = c(117L, 
    NA, 117L), Y = c(89L, NA, NA)), class = "data.frame", row.names = c(NA, 
-3L))

Upvotes: 2

M--
M--

Reputation: 28826

You should use | instead of &:

library(dplyr)

df1 %>%
  filter(!is.na(KeyPress) | !is.na(KPIndex) | !is.na(X) | !is.na(Y))

#   S.No MediaName KeyPress KPIndex     Type Secs   X  Y
# 1    1       Dat       NA       1 Fixation   18 117 89
# 2    3       Dat       NA       2 Fixation   23 117 NA

Upvotes: 2

Related Questions