To get the most frequent levels in a data frame

Question

I have the data frame like

now i need to get the output with most frequently occurred levels like

Could any one help me on this please??

Answer 1

We can use Mode function from here

 Mode <- function(x) {
  ux <- unique(x)
  ux[which.max(tabulate(match(x, ux)))]
}

and use it in any of the group by functions. One option is with dplyr

library(dplyr)
df %>%
   group_by(location) %>% 
   summarise(site = Mode(site))
#    location  site
#     <chr> <chr>
#1        a  site
#2        b  bang
#3        c  site

---

Or with base R

with(df, tapply(as.character(site), location, FUN = Mode))

data

 df<-data.frame(location=rep(letters[1:3], c(3, 1, 2)),
     site = c("site", "site", "bang", "bang", "site", "bang"), stringsAsFactors=FALSE)

Answer 2

You can do something like this with dplyr

set.seed(43)
df<-data.frame(location=sample(LETTERS[1:3],20,replace=TRUE),
               site=sample(c("bang","mys","hubl","dar"),20,replace=TRUE))

library(dplyr)
df%>%group_by(location,site)%>%summarize(Count=n())%>%arrange(desc(Count))%>%slice(1)%>%ungroup()%>%select(location,site)