20150211

library("plyr")
library("dplyr")
 
### create some sample data
set.seed(20)
dat <- data.frame(lake = c(rep("Muddy", 6), rep("Clear", 6)),
                  host = c(rep(c("a", "b"), 6)),
                  var = sample(c(0, 1, 2), 12, replace = TRUE),
                  stringsAsFactors = FALSE)
 
 
### dplyr solution
dat %>%
  group_by(lake, host) %>%
  summarize(prop = mean(var != 0))
 
 
### plyr solution
ddply(dat, .(lake, host), summarize, prop = mean(var != 0))


### base R solution
aggregate(var ~ lake + host, data = dat, function(x) mean(x != 0))


### to get proportions and numbers of observations at the same time with dplyr:
dat %>% 
  group_by(lake, host) %>%
  summarize(prop = mean(var != 0 ), 
            count = n())