Match opponent leads to team game logs and select variables for new, focused data table, as well as a summary table for each team.

opponent_team_lead_table_plus_summary.R

## set keys for the data table on game_slug and opponent_slug
team_game_logs <- setkey(team_game_logs, game_slug, opponent_slug)

## get opponent leads by matching on both key columns
team_game_logs$opp_biggest_lead <- team_game_logs[. (team_game_logs$game_slug, team_game_logs$opponent_slug)]$points_biggest_lead

## set keys for the data table on game_slug and opponent_slug
team_game_logs <- setkey(team_game_logs, game_slug, opponent_slug)

## get opponent leads by matching on both key columns
team_game_logs$opp_biggest_lead <- team_game_logs[. (team_game_logs$game_slug, team_game_logs$opponent_slug)]$points_biggest_lead

## select variables for new table
## rename so you don't lose original data
teamgame_point_leads <- select(team_game_logs, 
                               id, 
                               game_slug, 
                               team_slug, 
                               team_outcome, 
                               points_biggest_lead, 
                               opponent_slug, 
                               opp_biggest_lead, 
                               scoreline, 
                               score_differential, 
                               date)
## add final point differential
teamgame_point_leads$WL <- ifelse(teamgame_point_leads$team_outcome == "win", 1, -1)
teamgame_point_leads$final_point_diff <- (teamgame_point_leads$score_differential * teamgame_point_leads$WL)

## make a summary table for teamgame leads by team
## group records by team_slug
summary_teamgame_leads <- group_by(teamgame_point_leads, team_slug)

## select summary functions
summary_teamgame_leads <- summarise(summary_teamgame_leads,
	mean_biggest_lead = mean(points_biggest_lead, na.rm = TRUE),
	max_biggest_lead = max(points_biggest_lead, na.rm = TRUE), 
	median_biggest_lead = median(points_biggest_lead, na.rm = TRUE),
	opp_mean_biggest_lead = mean(opp_biggest_lead, na.rm = TRUE),
	opp_max_biggest_lead = max(opp_biggest_lead, na.rm = TRUE),
	opp_median_biggest_lead = median(opp_biggest_lead, na.rm = TRUE)
)