Merge pull request #26 from nflverse/clean_home_away

clean_homeaway

Project.toml

@@ -1,7 +1,7 @@
 name = "NFLData"
 uuid = "38e18452-fdda-4cae-b91e-088906595f57"
 authors = ["John Edwards"]
-version = "1.0.1"
+version = "1.1.1"
 
 [deps]
 Artifacts = "56f22d72-fd6d-98f1-02f0-08ddc0907c33"

docs/Manifest.toml

@@ -2,7 +2,7 @@
 
 julia_version = "1.10.5"
 manifest_format = "2.0"
-project_hash = "f97fbff8b171474864bc6f12303a9122440eaed6"
+project_hash = "cddd238c9b58118c59f6d32cb36945f8eb1002a3"
 
 [[deps.ANSIColoredPrinters]]
 git-tree-sha1 = "574baf8110975760d391c710b6341da1afa48d8c"

docs/Project.toml

@@ -1,3 +1,4 @@
 [deps]
+DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 NFLData = "38e18452-fdda-4cae-b91e-088906595f57"

docs/make.jl

@@ -1,4 +1,4 @@
-using Documenter, NFLData
+using Documenter, NFLData, DataFrames
 
 makedocs(
     sitename="NFLData.jl",

docs/src/helpers.md

@@ -6,6 +6,10 @@ Some helper functions are made available in `NFLData.jl` for use in loading and
 clean_player_names(player_name::String; lowercase::Bool = false, convert_lastfirst::Bool = true, use_name_database::Bool = true, convert_to_ascii::Bool = true)
 ```
 
+```@docs
+clean_homeaway(dataframe::AbstractDataFrame;invert=missing)
+```
+
 ```@docs
 most_recent_season(roster::Bool = false)
 ```

src/NFLData.jl

@@ -51,5 +51,6 @@ export get_current_week
 export nflverse_game_id
 export clean_team_abbrs
 export clean_player_names
+export clean_homeaway
 
 end

src/helpers.jl

@@ -10,6 +10,7 @@ export compute_labor_day
 export nflverse_game_id
 export clean_team_abbrs
 export clean_player_names
+export clean_homeaway
 
 function __init__()
     global team_abbr_mapping = CSV.read(joinpath(artifact"data","team_abbr_mapping.csv"),DataFrame)
@@ -101,7 +102,6 @@ julia> clean_player_names("Gordon Jr., Melvin", convert_lastfirst = true)
 "Melvin Gordon"
 ```
 """
-
 function clean_player_names(player_name::String; lowercase::Bool = false, convert_lastfirst::Bool = true, use_name_database::Bool = true, convert_to_ascii::Bool = true)
 
     player_name = strip(replace(player_name,r"\s+"=>" "))
@@ -158,4 +158,55 @@ function nflverse_game_id(season::Number,week::Number,away::String,home::String)
     return ids
 end
 
+"""
+    clean_homeaway(dataframe::AbstractDataFrame;invert = missing)
+
+Take a dataframe that is formatted with one record for a game between two teams and pivot it such that there exists two records per game, one for each team.
+
+Columns should be formatted such that any columns for data belonging to the home team are prefixed or suffixed with "home_*" and "*_home", likewise for away teams.
+
+Pass in a list of columns to `invert` to have these values multiplied by -1 before being returned to the new dataframe (such as margin of victory, which may be +7 for a home team and -7 for an away team in a given game).
+"""
+function clean_homeaway(dataframe::AbstractDataFrame;invert = missing)
+    home = deepcopy(dataframe)
+    away = deepcopy(dataframe)
+
+    rename!(home, replace.(names(home),r"^home_"=>"team_"))
+    rename!(home, replace.(names(home),r"^away_"=>"opponent_"))
+    rename!(home, replace.(names(home),r"_home$"=>""))
+    rename!(home, replace.(names(home),r"_away$"=>"_opponent"))
+    rename!(home, replace.(names(home),r"team_team"=>"team"))
+    rename!(home, replace.(names(home),r"opponent_team"=>"opponent"))
+
+    if "location" in names(home)
+        home.location = coalesce.(ifelse.(uppercase.(home.location) .== "NEUTRAL","neutral","home"),"home")
+    else
+        home.location .= "home"
+    end
+
+    rename!(away, replace.(names(away),r"^away_"=>"team_"))
+    rename!(away, replace.(names(away),r"^home_"=>"opponent_"))
+    rename!(away, replace.(names(away),r"_away$"=>""))
+    rename!(away, replace.(names(away),r"_home$"=>"_opponent"))
+    rename!(away, replace.(names(away),r"team_team"=>"team"))
+    rename!(away, replace.(names(away),r"opponent_team"=>"opponent"))
+
+    if "location" in names(away)
+        away.location = coalesce.(ifelse.(uppercase.(away.location) .== "NEUTRAL","neutral","away"),"away")
+    else
+        away.location .= "away"
+    end
+
+    if !ismissing(invert)
+        if !all(in.(invert, [names(away)]))
+            throw(DomainError(invert[.!in.(invert, [names(away)])],"Invalid cols passed to `invert`!"))
+        end
+        for col in invert
+            away[:,col] = -1 .* away[:,col]
+        end
+    end
+
+    return(vcat(home, away))
+end
+
 end