twitter - Remove "semi-duplicate" rows in R -

i have dataset looks this:

  text                 id     screenname retweetcount isretweet retweeted longitude latitude 1    xx 778980737861062656  0504traveller            0     false     false      <na>     <na> 2    xx 778967536167559168       iz_azman            0     false     false      <na>     <na> 3    yy 778962265298960384       iz_azman            0     false     false      <na>     <na> 4    yy 778954988122939392 travelindtoday            2     false     false      <na>     <na> 5    zz 778948691969224705           umtn            2     false     false      <na>     <na> 6    zz 778942095843135493     flyinsider            0     false     false      <na>     <na> 

these tweets package twittr in r. tweets have same text different retweetcount. want keep unique tweets (by text), keeping highest retweetcount amongst duplicates. (in case above, tweets 1, 4, , 5.)

how do that?

you can dplyr

library(dplyr) df %>%   group_by(text) %>%   slice(which.max(retweetcount))  #text           id     screenname retweetcount isretweet retweeted longitude latitude #(fctr)        (dbl)         (fctr)        (int)     (lgl)     (lgl)    (fctr)   (fctr) #1  xx      7.789807e+17  0504traveller       0     false     false      <na>     <na> #4  yy      7.789550e+17 travelindtoday       2     false     false      <na>     <na> #5  zz      7.789487e+17           umt        2     false     false      <na>     <na> 

---

another approach in base r using aveand order :

df[ave(df$retweetcount,df$text, fun = function(x) order(x, decreasing = t)) == 1, ]  # text           id     screenname retweetcount isretweet retweeted longitude latitude #1  xx   7.789807e+17  0504traveller            0     false     false      <na>     <na> #4  yy   7.789550e+17 travelindtoday            2     false     false      <na>     <na> #5  zz   7.789487e+17           umtn            2     false     false      <na>     <na>