CODEWITHSUNDEEP
rdplyr
datakritter
datakritter

Reputation: 600

Can't get dplyr::ungroup to ungroup

I'm sure I'm missing something obvious here. I want to add rownumbers to a summarized table, but dplyr cycles/repeats the rownumbers per my earlier grouping. I thought I had ungrouped it, but clearly I'm missing something. I want each line in indexed to have a unique rownumber.

#Make sample data
    Species <- c("A", "B", "C","D")
    Animal <- c(1:100)
    Day <-c(1:10)
    P <- sample(1:100, 300, replace = TRUE)
    F <- sample(1:100, 300, replace = TRUE)
    C <- sample(1:100, 300, replace = TRUE)
    df <- data.frame(Species,Animal,Day, P, F, C)

 #Summarize by columns
    by_day <- df %>%
      group_by(Species, Animal, Day) %>%
      summarize(ptot = sum(P), ftot = sum(F), ctot = sum(C))

 #Here's where I suspect the problem lies    
    ungroup(by_day)

 #This line produces repeating id numbers, where as I want each line to have a unique one.
    indexed <- mutate(by_day, id = row_number())

Upvotes: 1

Views: 6653

Answers (2)

deca
deca

Reputation: 740

If you simply reformat a grouped data frame as data frame via as.data.frame(), it is also automatically ungrouped.

Upvotes: 0

akuiper
akuiper

Reputation: 214957

It seems you assume ungroup will modify by_day in place, which it doesn't; You need to make sure you pass an ungrouped data frame to mutate:

mutate(ungroup(by_day), id = row_number())

# A tibble: 100 x 7
#   Species Animal   Day  ptot  ftot  ctot    id
#    <fctr>  <int> <int> <int> <int> <int> <int>
# 1       A      1     1   266   262    45     1
# 2       A      5     5    84   201   159     2
# 3       A      9     9   141   149   244     3
# 4       A     13     3    94   142   157     4
# 5       A     17     7   188   138   142     5

1) before ungroup:

by_day
# A tibble: 100 x 6
# Groups:   Species, Animal [?]        <<<<<<<<<<<<<<<<<<<<
#   Species Animal   Day  ptot  ftot  ctot

2) ungroup returns a data frame that is not grouped:

ungroup(by_day)
# A tibble: 100 x 6
#   Species Animal   Day  ptot  ftot  ctot

3) but it doesn't modify by_day, it's still grouped

by_day
# A tibble: 100 x 6
# Groups:   Species, Animal [?]        <<<<<<<<<<<<<<<<<<<<
#   Species Animal   Day  ptot  ftot  ctot

Upvotes: 6

Related Questions