I have a dummy dataframe with four columns.

df <- data.frame(City = c("A","A","A","B","B","B","B"),
                 Name=c("Jon", "Bill","Bill", "Maria", "Ben", "Tina",'Tina'),
                 Age = c(23, 41, 32, 58, 26, 12, 15),
                 Eye_color=c("Blue","Blue", "Brown", "Brown", "Blue", "Blue","Brown"))

  City  Name Age Eye_color
1    A   Jon  23      Blue
2    A  Bill  41      Blue
3    A  Bill  32     Brown
4    B Maria  58     Brown
5    B   Ben  26      Blue
6    B  Tina  12      Blue
7    B  Tina  15     Brown

I want to remove duplicates in Names (Bill and Tina) with two different cases:

First case: group by City and remove duplicates in Names keeping the Blue eyed only. Result 1 should look like this:

  City  Name Age Eye_color
1    A   Jon  23      Blue
2    A  Bill  41      Blue
3    B Maria  58     Brown
4    B   Ben  26      Blue
5    B  Tina  12      Blue

Second case: I want to specify that if the city is A, between the duplicates in Names keep Blue eye, if the City is B between the duplicates in Name keep the Brown eye.

Result 2 should look like this:

  City  Name Age Eye_color
1    A   Jon  23      Blue
2    A  Bill  41      Blue
3    B Maria  58     Brown
4    B   Ben  26      Blue
5    B  Tina  15     Brown

Thanks for the help!

CodePudding user response：

Here is one possibility using filter and dplyr:

First we filter for Eye_color == Blue but only if one row contains ´Blue`.

df %>%
  group_by(Name) %>%
  filter(if (any(Eye_color == "Blue")) Eye_color == "Blue" else TRUE) %>%
  ungroup()

In the second case we use if_else in the filter statement:

df %>%
  filter(if_else(Name == "Bill", Eye_color == "Blue", if_else(Name == "Tina", Eye_color == "Brown", TRUE)))

Update

For the new dataset you can use the same code for part 1. For part 2 simply replace the logical statements inside if_else:

df %>%
  filter(if_else(City == "A", Eye_color == "Blue", if_else(City == "B", Eye_color == "Brown", TRUE)))

CodePudding user response：

You can use this code:

df1 <- df %>%group_by(Name) %>% filter(Eye_color == "Blue")
df2 <- df %>% filter(if_else(Name == "Bill", Eye_color == "Blue", if_else(Name == "Tina", Eye_color == "Brown", TRUE)))