spread does not work on grouped_df

[tpapp]

Apparently spread does not work on grouped_df objects. Toy example:

library(dplyr)                          # from github, 2014 Oct 31
library(tidyr)                          # from github, 2014 Oct 31

## generate data
household <- rep(1:500, times = sample(5, 500, replace = TRUE))
df <- data.frame(household = household,
                 day = sample(3, length(household), replace = TRUE))

## count day indexes for each household
df_days <- summarize(group_by(df, household, day), day_count = n())
## spread, converting to data frame -- works
spread(as.data.frame(df_days), day, day_count)
## spread without conversion -- does not work
spread(df_days, day, day_count)

where the last line fails with

Error in `[.grouped_df`(data, key_col) : 
  cannot group, grouping variables 'household' not included

Maybe I misunderstood something, but I thought grouped_df could be used like data frame.

[hadley]

This should be fixed in the dev version of dplyr