Aggregating dataframe by occurrences of 3 columns in R

Question

Here is my data frame:

       Col1      Col2          Col3       Day       
 1       12         2  14031100000417017  70
 2       10         3  14031100000537996   6
 3        7         4  14031100000636479  11
 4       12         2  14031100001124050  70
 5       12         2  14031100001712518  70
 6       10         3  14031100001920801   6

How can I create a new data frame that shows me how many times Col1, Col2 and Day occur with the same values? So in this example I would get:

       Col1      Col2          Col3       Day   Count    
 1      12         2   14031100000417017  70      3
 2      10         3   14031100000537996   6      2
 3       7         4   14031100000537996  11      1

Answer 1

This is a good case for using the dplyr package, described in detail here.

library(dplyr)
my.dataframe %.% group_by(Col1, Col2, Day) %.% summarise(Count=n())

This can also be written as

summarise(group_by(my.dataframe, Col1, Col2, Day), Count=n())

Answer 2

In base R, you can just use ave to create a "count" column. You don't really specify how you want to deal with "Col3", so I've left that out:

mydf$count <- ave(rep(1, nrow(mydf)), 
                  mydf[c("Col1", "Col2", "Day")], 
                  FUN = length)

mydf
#   Col1 Col2        Col3 Day count
# 1   12    2 1.40311e+16  70     3
# 2   10    3 1.40311e+16   6     2
# 3    7    4 1.40311e+16  11     1
# 4   12    2 1.40311e+16  70     3
# 5   12    2 1.40311e+16  70     3
# 6   10    3 1.40311e+16   6     2

unique(mydf[c("Col1", "Col2", "Day", "count")])
#   Col1 Col2 Day count
# 1   12    2  70     3
# 2   10    3   6     2
# 3    7    4  11     1

Answer 3

And to round it off, here's a data.table solution:

library(data.table)
mydt <- data.table(mydf)
mydt[, length(Col3), list(Col1, Col2, Day)]

#    Col1 Col2 Day V1
# 1:   12    2  70  3
# 2:   10    3   6  2
# 3:    7    4  11  1

The third argument is a list of grouping factors, while the second arg is the function that you'd like to apply to each group.

Answer 4

Or using plyr:

require(plyr)
ddply(df, .(Col1, Col2, Day), summarize, cnt = length(Col1))

or aggregate:

aggregate(Col3 ~ Col1 + Col2 + Day, df, length)

  Col1 Col2 Day Col3

1   10    3   6    2
2    7    4  11    1
3   12    2  70    3