A.10 Gluing data.frames using bind_rows

Suppose we want to find average score by team. We’ll have to rearrange the data a little since teams can appear in both the home and away column. Ideally, we would have one column with the team name, and one column with the score. Each game would then have two rows, one for the away team and one for the home team. We’ll create two data.frames, one for home teams and one for away teams, and then bind them together using bind_rows.

Code

da = d %>% select(date, away, ascore, home, hscore, gid) %>% mutate(ha = 'away')
dh = d %>% select(date, home, hscore, away, ascore, gid) %>% mutate(ha = 'home')
colnames(da) = c('date', 'team', 'score',  'opp', 'opp.score', 'gid', 'ha')
colnames(dh) = c('date', 'team', 'score',  'opp', 'opp.score', 'gid', 'ha')
dd = bind_rows(da, dh) %>% 
  arrange(date, gid)
head(dd)

        date team score opp opp.score      gid   ha
1 2021-10-19  BKN   104 MIL       127 22100001 away
2 2021-10-19  MIL   127 BKN       104 22100001 home
3 2021-10-19  GSW   121 LAL       114 22100002 away
4 2021-10-19  LAL   114 GSW       121 22100002 home
5 2021-10-20  IND   122 CHA       123 22100003 away
6 2021-10-20  CHA   123 IND       122 22100003 home

We see the same games, teams, and scores at the top, but now we have two rows for each game. In this case, I put the column names in the select and colnames rows on the same row so that it is easier to compare and contrast the first and second lines, and third and fourth lines.