CODEWITHSUNDEEP
rdataframereplace
useryk
useryk

Reputation: 309

Replace values from multiple columns based on value from adjacent column

    # Create a data frame
        > df <- data.frame(a = rnorm(7), b = rnorm(7), c = rnorm(7), threshold = rnorm(7))
        > df <- round(abs(df), 2)
        > 
        > df
             a    b    c threshold
        1 1.17 0.27 1.26      0.19
        2 1.41 1.57 1.23      0.97
        3 0.16 0.11 0.35      1.34
        4 0.03 0.04 0.10      1.50
        5 0.23 1.10 2.68      0.45
        6 0.99 1.36 0.17      0.30
        7 0.28 0.68 1.22      0.56
        > 
        >
    # Replace values in columns a, b, and c with NA if > value in threshold
        > df[1:3][df[1:3] > df[4]] <- "NA"
        Error in Ops.data.frame(df[1:3], df[4]) : 
          ‘>’ only defined for equally-sized data frames

There could be some obvious solutions that I am incapable of producing. The intent is to replace values in columns "a", "b", and "c" with NA if the value is larger than that in "threshold". And I need to do that row-by-row.

If I had done it right, the df would look like this:

         a    b    c threshold
    1   NA   NA   NA      0.19
    2   NA   NA   NA      0.97
    3 0.16 0.11 0.35      1.34
    4 0.03 0.04 0.10      1.50
    5 0.23   NA   NA      0.45
    6   NA   NA 0.17      0.30
    7 0.28   NA   NA      0.56

I had also tried the apply() approach but to no avail. Can you help, please??

Upvotes: 2

Views: 370

Answers (4)

markus
markus

Reputation: 26343

The problem with your code was the usage of df[4] instead of df[, 4]. The difference is that df[4] returns a data.frame with one column and df[, 4] returns a vector.

That's why

df[1:3] > df[4]

returns

error in Ops.data.frame(df[1:3], df[4]) : ‘>’ only defined for equally-sized data frames

While this works as expected

df[1:3][df[1:3] > df[, 4]] <- NA
df
#     a    b    c threshold
#1 0.63 0.74   NA      0.78
#2   NA   NA 0.04      0.07
#3 0.84 0.31 0.02      1.99
#4   NA   NA   NA      0.62
#5   NA   NA   NA      0.06
#6   NA   NA   NA      0.16
#7 0.49   NA 0.92      1.47

data

set.seed(1)
df <- data.frame(a = rnorm(7), b = rnorm(7), c = rnorm(7), threshold = rnorm(7))
df <- round(abs(df), 2)

Upvotes: 1

Arun kumar mahesh
Arun kumar mahesh

Reputation: 2359

You can use apply function across dataframe

df[,c(1:3)]<- apply(df[,c(1:3),drop=F], 2, function(x){ ifelse(x>df[,4],NA,x)})

Upvotes: 2

Sonny
Sonny

Reputation: 3183

You should use dplyr for most of such use cases. One way below:

> set.seed(10)
> df <- data.frame(a = rnorm(7), b = rnorm(7), c = rnorm(7), threshold = rnorm(7))
> df <- round(abs(df), 2)
> df
     a    b    c threshold
1 0.02 0.36 0.74      2.19
2 0.18 1.63 0.09      0.67
3 1.37 0.26 0.95      2.12
4 0.60 1.10 0.20      1.27
5 0.29 0.76 0.93      0.37
6 0.39 0.24 0.48      0.69
7 1.21 0.99 0.60      0.87
> 
> df %>%
+   mutate_at(vars(a:c), ~ifelse(.x > df$threshold, NA, .x))
     a    b    c threshold
1 0.02 0.36 0.74      2.19
2 0.18   NA 0.09      0.67
3 1.37 0.26 0.95      2.12
4 0.60 1.10 0.20      1.27
5 0.29   NA   NA      0.37
6 0.39 0.24 0.48      0.69
7   NA   NA 0.60      0.87

Upvotes: 3

morgan121
morgan121

Reputation: 2253

You can use a for-loop like this:

for(i in 1:(ncol(df)-1)){ 
  df[, i] <- ifelse(df[, i] > df[, 4], NA, df[, i])
}

Upvotes: 0

Related Questions