How to choose the best in each year of a column in R?

Asked

Viewed 29 times

1

I have a table with the efficiency of each player, the year of the season and the name of the player. How can I get you to give me a final table with the best player (highest efficiency) each year? Giving me the player’s name and its efficiency and year on a table

r

1 answer

2

Using the package dplyr or data.table it is possible to do this.

  • Creating data:
# dados ficticios
set.seed(1)
df <- data.frame(ano = rep(2015:2018, each = 5),
                 nome = rep(letters[1:4], 5),
                 eficiencia = rnorm(n = 20, mean = 0.6, sd = 0.1))
df

> df
    ano nome eficiencia
1  2015    a  0.5373546
2  2015    b  0.6183643
3  2015    c  0.5164371
. . .
18 2018    b  0.6943836
19 2018    c  0.6821221
20 2018    d  0.6593901
  • Applying the package dplyr with top_n:
library(dplyr)

df2 <- df %>% 
  dplyr::group_by(ano) %>% # agrupar por ano
  dplyr::top_n(1)          # apenas o melhor (1) em cada ano
df2

> df2
# A tibble: 4 x 3
# Groups:   ano [4]
    ano nome  eficiencia
  <int> <fct>      <dbl>
1  2015 d          0.760
2  2016 d          0.674
3  2017 c          0.751
4  2018 b          0.694

Or with slice:

df3 <- df %>% 
  dplyr::group_by(ano) %>% 
  dplyr::slice(which.max(eficiencia))
df3

> df3
# A tibble: 4 x 3
# Groups:   ano [4]
    ano nome  eficiencia
  <int> <fct>      <dbl>
1  2015 d          0.760
2  2016 d          0.674
3  2017 c          0.751
4  2018 b          0.694
  • Applying the package data.table:
library(data.table)
dft <- data.table::data.table(df)

df4 <- dft[dft[, .I[eficiencia == max(eficiencia)], by = ano]$V1]
df4

> df4
    ano nome eficiencia
1: 2015    d  0.7595281
2: 2016    d  0.6738325
3: 2017    c  0.7511781
4: 2018    b  0.6943836

Or with:

df5 <- data.table::setDT(dft)[, .SD[which.max(eficiencia)], by = ano]
df5

> df5
    ano nome eficiencia
1: 2015    d  0.7595281
2: 2016    d  0.6738325
3: 2017    c  0.7511781
4: 2018    b  0.6943836

Browser other questions tagged r

You are not signed in. Login or sign up in order to post.