CODEWITHSUNDEEP
r
Hack-R
Hack-R

Reputation: 23200

Remove subsequent elements with the same substring

I have an n by 2 object containing variable names in the 1st column and numeric values (scores) in the 2nd column:

 data      <- data.frame(matrix(nrow = 20, ncol = 2))
 data[, 2] <- 1:20
 data[, 1] <- c("example_a_1", "example_a_2", "example_a_3",
                "example_b_1", "example_c_1", "example_d_1",
                "example_d_2", "example_d_3", "example_f_1",
                "example_g_1", "example_g_2", "example_h_1",
                "example_i_1", "example_l_1", "example_o_1",
                "example_j_1", "example_m_1", "example_p_1",
                "example_k_1", "example_n_1")
data
            X1 X2
1  example_a_1  1
2  example_a_2  2
3  example_a_3  3
4  example_b_1  4
5  example_c_1  5
6  example_d_1  6
7  example_d_2  7
8  example_d_3  8
9  example_f_1  9
10 example_g_1 10
11 example_g_2 11
12 example_h_1 12
13 example_i_1 13
14 example_l_1 14
15 example_o_1 15
16 example_j_1 16
17 example_m_1 17
18 example_p_1 18
19 example_k_1 19
20 example_n_1 20

I don't want this object to contain similar variables -- if a variable name has the same first 9 characters (in this example) as another, then it's repetitious. In those cases I only want to keep the first of the similarly named variables.

I can get a list of which variable names are repetitious like this:

 rep <- as.data.frame(table(substr(data[,1], 1, 9)))
 rep <- rep[rep[, 2] > 1, ]
 rep
       Var1 Freq
1 example_a    3
4 example_d    3
6 example_g    2

and thus identify them in a for loop or other conditional:

for(i in 1:nrow(data)){
  if(substr(data[i, 1], 1, 9) %in% rep[,1])){
    # What goes here?
    #   or what's another approach?
  }
}

However, I'm not sure what logic I can use to remove the rows with repetitious names?

The final object should look like this:

data
           X1 X2
1 example_a_1  1
2 example_b_1  4
3 example_c_1  5
4 example_d_1  6
5 example_f_1  9
6 example_g_1 10
7 example_h_1 12
8 example_i_1 13
9 example_l_1 14
10 example_o_1 15
11 example_j_1 16
12 example_m_1 17
13 example_p_1 18
14 example_k_1 19
15 example_n_1 20

Upvotes: 2

Views: 57

Answers (3)

zx8754
zx8754

Reputation: 56169

Using dplyr:

library(dplyr)
data  <- data %>% 
            group_by(my9=substr(X1,1,9)) %>% 
            filter(row_number(my9)==1) %>% 
            select(-my9)

Upvotes: 2

Michal
Michal

Reputation: 1905

I would create a column with the shortened name and aggregate on that column:

data$short <- substr(data[,1], 1, 9)
agg <- aggregate(data$X2~data$short, FUN=min)

I used min because you seem to be interested in the smallest score for each repetitive name

Upvotes: 2

Robert Hijmans
Robert Hijmans

Reputation: 47146

You can use duplicated

short <- substr(data[,1], 1, 9)
i <- duplicated(short)
data <- data[!i , ]

Upvotes: 5

Related Questions