如何删除R中存在重复并满足另一个条件的行？

Question

I don't know if this may be a too specific question, but I'm looking to remove rows which have duplicates in one column, and meet a condition.

To be specific, I want to delete one of the duplicate observations in the column "host_id" (numeric), for which the value in the column "reviews_per_month" (numeric) is the lowest.

In other words, as described in my report: " Since one host can have multiple listings, hosts ids that appear more than one time will be filtered. The listing of this host's id which has the most reviews per month is used for analysis ".

I've tried many things using duplicated(), filter(), ifelse(), casewhen(), etc, but it doesn't seem to work. Does anyone know how to get started? Thanks in advance!

Answer 1

We can use slice_max . Grouped by 'host_id', slice the row where the reviews_per_month is the max

library(dplyr)
df %>%
   group_by(host_id) %>%
   slice_max(reviews_per_month)

Or if it is to remove the min observation alone

df %>%
   group_by(host_id) %>%
   filter(reviews_per_month != min(reviews_per_month, na.rm = TRUE))

Answer 2

Try this:

df <- data.frame(x = c("a", "a", "b", "b"), y = c(1, 2, 2, 1)) # Test data

library(dplyr)

df %>% 
  distinct(x, .keep_all = T)
# Wrong!

df %>% 
  arrange(-y) %>% 
  distinct(x, .keep_all = T)
# This is how you want to have it

To be a bit more verbose: You just want one entry in your host_id variable ( x in the example above), so you'll want to use distinct() . But distinct() keeps just the first observation of the variable that is fed into distinct (in your case: host_id ), so you have to sort the data in decreasing fashion first. I use arrange(-y) in my example, you should replace y by reviews_per_month .