碱基R：

dt[dt$City %in% names(sort(table(dt$City), decreasing = TRUE)[1:2]), ]

# version by @M-- using tail
dt[ dt$City %in% names( tail( table( dt$City) , 2) ) , ]

 ID     City Random_Info_To_Keep
1   1 New York                   1
2   2 New York                   1
3   3       LA                   1
4   4       LA                   1
5   5       LA                   1
8   8 New York                   1
9   9       LA                   1
10 10 New York                   1

您可以首先计算每个城市在数据集中出现的次数，然后过滤数据中出现次数最多的两个城市。以下是tidyverse解决方案：

library(tidyverse)

df %>%
  add_count(City, name = "helper_count") %>% # adding the frequency with which each city occurs
  slice_max(order_by = helper_count, n = 2) %>% # two most common
  select(-helper_count)