How to fill whole pie chart in facet wraps without using scale = free

further to a previous question How to subset or aggregate large amounts of data so I can make separate pie charts

I have managed to produce the below pie charts, which is great, however does anyone know a way that I may have the pie charts filled up according to Trip_Set? i.e., most of the Trip_Sets have just one male or one female so they should be totally pink or blue. Because one has 3 (271_6), it's taken that (the highest) as default so the other pies have the empty space. I know that if I were doing this in bar charts I could add in scales = "free"but this doesn't work with with a non-cartesian coord system...

Here is my data:

Trip_Set   sex
119_4      hembra
119_4      hembra
119_7      hembra
161_7      macho
193_8      hembra
255_7      macho
271_6      hembra
271_6      macho
271_6      hembra
328_7      hembra
403_3      hembra
428_2      hembra
655_4      hembra

Here is my code:

pie <- ggplot(dframe2, aes(x=1, y='Trip_Set', fill=sex)) +
  geom_bar(width = 1, stat="identity") +
  ggtitle("Sex ratio per line") +
  coord_polar(theta='y') +
  facet_wrap(~Trip_Set, ncol = 5) 

See if this is what you're looking for:

# transform the data
library(dplyr)
dframe2_new = dframe2 %>% group_by(Trip_Set, sex) %>% tally()
# to plot
ggplot(data=dframe2_new, aes(x=factor(1), y=n, fill=sex)) +
  geom_bar(stat="identity", position=position_fill(width=1)) +
  coord_polar(theta="y") +
  facet_wrap(~Trip_Set, ncol=5)

If you don't want coordinates etc., you could try the code below

# to plot
ggplot(data=dframe2_new, aes(x=factor(1), y=n, fill=sex)) +
  geom_bar(stat="identity", position=position_fill(width=1)) +
  coord_polar(theta="y") +
  facet_wrap(~Trip_Set, ncol=5) + 
  theme(axis.text = element_blank(),
        axis.ticks = element_blank(),
        panel.grid  = element_blank()) +
  xlab("") +
  ylab("Trip_Set")