I have many rosters of people with statistics that I made into a data.frame teamroster. Only problem is, some people have duplicate names, and don't belong on a roster (because they have a different team name. Look at case Matt Duffy in teamroster below). I want to systematically remove all names and entries that do not have the same team name on the roster.
Here is my raw data.frame:
teamroster
Name Team G PA
1 Denard Span Giants 30 135
2 Joe Panik Giants 25 107
3 Matt Duffy Giants 31 127
4 Matt Duffy Astros 3 3
5 Buster Posey Giants 27 108
The solution code will recognize that Matt Duffy is on a different team, as seen by the Team column, and remove him because he is on Team = Astros. This is what I want the resulting data frame to look like:
finishedteamroster
Name Team G PA
1 Denard Span Giants 30 135
2 Joe Panik Giants 25 107
3 Matt Duffy Giants 31 127
4 Buster Posey Giants 27 108
You could tabulate the team names then take the maximum of the tabulation. Note that I used which.max() for its side-effect of keeping the table names.
idx <- with(df, Team == names(which.max(table(Team))))
df[idx, ]
# Name Team G PA
# 1 Denard Span Giants 30 135
# 2 Joe Panik Giants 25 107
# 3 Matt Duffy Giants 31 127
# 5 Buster Posey Giants 27 108
Data:
df <- structure(list(Name = structure(c(2L, 3L, 4L, 4L, 1L), .Label = c("Buster Posey",
"Denard Span", "Joe Panik", "Matt Duffy"), class = "factor"),
Team = structure(c(2L, 2L, 2L, 1L, 2L), .Label = c("Astros",
"Giants"), class = "factor"), G = c(30L, 25L, 31L, 3L, 27L
), PA = c(135L, 107L, 127L, 3L, 108L)), .Names = c("Name",
"Team", "G", "PA"), class = "data.frame", row.names = c("1",
"2", "3", "4", "5"))