This is part of a data frame I loaded from the internet using readHTMLtable:
head(tt,59)
year sport event athlete_id medal
1 1896 Track & Field 100m Men BURKETOM01 GOLD
2 1896 Track & Field 100m Men HOFMAFRI01 SILVER
3 1896 Track & Field 100m Men LANEFRA01 BRONZE
4 1896 Track & Field 100m Men SZOKOALA01 BRONZE
5 1896 Track & Field 400m Men BURKETOM01 GOLD
6 1896 Track & Field 400m Men JAMISHER01 SILVER
7 1896 Track & Field 400m Men GMELICHA01 BRONZE
8 1896 Track & Field 800m Men FLACKTED01 GOLD
9 1896 Track & Field 800m Men D<C1>NIN<C1>N01 SILVER
10 1896 Track & Field 800m Men GOLEMDEM01 BRONZE
11 1896 Track & Field 1500m Men FLACKTED01 GOLD
12 1896 Track & Field 1500m Men BLAKEART01 SILVER
13 1896 Track & Field 1500m Men LERMUALB01 BRONZE
14 1896 Track & Field Marathon Men LOUISSPI01 GOLD
15 1896 Track & Field Marathon Men VASILCHA01 SILVER
16 1896 Track & Field Marathon Men KELLNGYU01 BRONZE
17 1896 Track & Field 110m Hurdles Men CURTITOM01 GOLD
18 1896 Track & Field 110m Hurdles Men GOULDGRA01 SILVER
19 1896 Track & Field High Jump Men CLARKELL01 GOLD
20 1896 Track & Field High Jump Men CONNOJAM01 SILVER
21 1896 Track & Field High Jump Men GARREBOB01 SILVER
22 1896 Track & Field Pole Vault Men HOYTBIL01 GOLD
23 1896 Track & Field Pole Vault Men TYLERALB01 SILVER
24 1896 Track & Field Pole Vault Men THEODIOA01 BRONZE
25 1896 Track & Field Pole Vault Men DAMASEVA01 BRONZE
26 1896 Track & Field Long Jump Men CLARKELL01 GOLD
27 1896 Track & Field Long Jump Men GARREBOB01 SILVER
28 1896 Track & Field Long Jump Men CONNOJAM01 BRONZE
29 1896 Track & Field Triple Jump Men CONNOJAM01 GOLD
30 1896 Track & Field Triple Jump Men TUFF<C8>ALE01 SILVER
31 1896 Track & Field Triple Jump Men PERSAIOA01 BRONZE
32 1896 Track & Field Shot Put Men GARREBOB01 GOLD
33 1896 Track & Field Shot Put Men GOUSKMIL01 SILVER
34 1896 Track & Field Shot Put Men PAPASGEO01 BRONZE
35 1896 Track & Field Discus Throw Men GARREBOB01 GOLD
36 1896 Track & Field Discus Throw Men PARASPAN01 SILVER
37 1896 Track & Field Discus Throw Men VERSISOT01 BRONZE
38 1896 Cycling 2000m Sprint (Scratch) Men MASSOPAU01 GOLD
39 1896 Cycling 2000m Sprint (Scratch) Men NIKOLSTA01 SILVER
40 1896 Cycling 2000m Sprint (Scratch) Men FLAMEL<C9>O01 BRONZE
41 1896 Cycling Individual Road Race Men KONSTARI01 GOLD
42 1896 Cycling Individual Road Race Men GOEDRAUG01 SILVER
43 1896 Cycling Individual Road Race Men BATTEEDW01 BRONZE
44 1896 Cycling One-Lap Race MASSOPAU01 GOLD
45 1896 Cycling One-Lap Race NIKOLSTA01 SILVER
46 1896 Cycling One-Lap Race SCHMAADO01 BRONZE
47 1896 Cycling 10km Track Race MASSOPAU01 GOLD
48 1896 Cycling 10km Track Race FLAMEL<C9>O01 SILVER
49 1896 Cycling 10km Track Race SCHMAADO01 BRONZE
50 1896 Cycling 100km Track Race FLAMEL<C9>O01 GOLD
51 1896 Cycling 100km Track Race KOLETGEO01 SILVER
52 1896 Cycling 12-Hour Race SCHMAADO01 GOLD
53 1896 Cycling 12-Hour Race KEEPIFRA01 SILVER
54 1896 Fencing Foil, Individual GRAVEEUG01 GOLD
55 1896 Fencing Foil, Individual CALLOHEN01 SILVER
56 1896 Fencing Foil, Individual PIERRPER01 BRONZE
57 1896 Fencing Sabre, Individual GEORGIOA01 GOLD
58 1896 Fencing Sabre, Individual KARAKTEL01 SILVER
59 1896 Fencing Sabre, Individual NIELSHOL01 BRONZE
As you can see the variable sport is a factor. When I check the levels this is what I get:
levels(tt$sport)
[1] "Cycling" "Fencing" "Gymnastics" "Shooting" "Swimming" "Tennis"
[7] "Track & Field" "Weightlifting" "Wrestling
For some reason the order in which the levels appear does not match the order in the data frame. I am looking for a way in which using levels function will give me a list of the levels organized according to the first appearance, something like that:
levels(medals.df$tt)
[1] "Track & Field" "Cycling" "Fencing" "Gymnastics" "Shooting" "Swimming"
[7] "Tennis" "Weightlifting" "Wrestling"
Now another thing to keep in mind is that the column sport is not in a "block design", meaning the first 59 rows have all the same values adjacent but it is not like this throughout the entire data frame.
Note that I had to tweak your dataset so that all the levels you list appear, and do so in the order you specified. From there, I wrote a simple function that outputs the levels in the order they appear in the dataset. The key is to use which (which lists the row numbers of observations that match a criterion), min (which selects the lowest value), and order (which tells you the order to use to go from the lowest to the highest).
d <- read.table(text="rn year sport event athlete_id medal
1 1896 'Track & Field' '100m Men' 'BURKETOM01' 'GOLD'
53 1896 'Cycling' '12-Hour Race' 'KEEPIFRA01' 'SILVER'
54 1896 'Fencing' 'Foil, Individual' 'GRAVEEUG01' 'GOLD'
55 1896 'Gymnastics' 'Foil, Individual' 'CALLOHEN01' 'SILVER'
56 1896 'Shooting' 'Foil, Individual' 'PIERRPER01' 'BRONZE'
57 1896 'Swimming' 'Sabre, Individual' 'GEORGIOA01' 'GOLD'
58 1896 'Tennis' 'Sabre, Individual' 'KARAKTEL01' 'SILVER'
58 1896 'Weightlifting' 'Sabre, Individual' 'KARAKTEL01' 'SILVER'
59 1896 'Wrestling' 'Sabre, Individual' 'NIELSHOL01' 'BRONZE'",
header=T)
levels(d$sport)
# [1] "Cycling" "Fencing" "Gymnastics" "Shooting"
# [5] "Swimming" "Tennis" "Track & Field" "Weightlifting"
# [9] "Wrestling"
level.order <- function(var){
l <- levels(var)
o <- c()
for(i in 1:length(l)){
o[i] <- min(which(var==l[i]))
}
return(l[order(o)])
}
level.order(d$sport)
# [1] "Track & Field" "Cycling" "Fencing" "Gymnastics"
# [5] "Shooting" "Swimming" "Tennis" "Weightlifting"
# [9] "Wrestling"
From here, if you wanted to change the default ordering (alphabetical) to the order the levels show up in the dataset, you would use factor. Consider:
levels(d$sport)
# [1] "Cycling" "Fencing" "Gymnastics" "Shooting"
# [5] "Swimming" "Tennis" "Track & Field" "Weightlifting"
# [9] "Wrestling"
d$sport <- factor(d$sport, levels=level.order(d$sport))
levels(d$sport)
# [1] "Track & Field" "Cycling" "Fencing" "Gymnastics"
# [5] "Shooting" "Swimming" "Tennis" "Weightlifting"
# [9] "Wrestling"