# There is a lot more to this question than you might initially realise... 

# If you think the answer is that the youngest director was 11 and the oldest
# director was 103 then you are wrong on both counts. 

# Maybe go back and think a bit more then try again before scrolling down to the
# answer below.


















# What I expect you to get to first, by recognising that where there are missing
# values we need to use na.rm=TRUE in the summary functions, and that a little
# bit of translation from human to R is needed: "youngest"=min(age) and
# "oldest"=max(age). You could also have added a line to filter() to remove the missing
# values out that way instead of the na.rm=TRUE.

imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
      summarise(Youngest=min(DirectorReleaseAge, na.rm=TRUE),  Oldest=max(DirectorReleaseAge, na.rm=TRUE))
      
# Hopefully some of you might start looking at those numbers and questioning these results though!
# And before going any further it is worth considering that around 24% of films are directed by people with an unknown birth year. So there are pretty decent chances that among that 24% there might be someone older or younger.
# But that is not the main issue though...

# Let's look at the oldest first, as 103 is provably an incorrect answer within R.

imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(DirectorReleaseAge==103)

# Orson Welles died in 1985. So don't think we can say with any validity that he
# was the "oldest" director for a film released 33 years after his death! We
# should either set posthumous releases to have a missing value for age or we
# should exclude them from the analysis. Let's go with the second option as this
# is a little bit cleaner to include in a piped series of commands. Now there
# may still be issues. NA for "deathYear" could mean that the director is still
# alive or it could mean that the deathYear is unknown. And someone could have died
# before their movie was released, but in the same year. So there may still be
# inconsistencies in our age calculations, but it is at least better than it was
# before

imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(year<=deathYear | is.na(deathYear)) %>%
      summarise(Youngest=min(DirectorReleaseAge, na.rm=TRUE),  Oldest=max(DirectorReleaseAge, na.rm=TRUE))

imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(year<=deathYear | is.na(deathYear)) %>%
    filter(DirectorReleaseAge==102)
      
# This one seems to be legit! https://en.wikipedia.org/wiki/Manoel_de_Oliveira
# "Beginning in the late 1980s he was one of the most prolific working film
# directors and made an average of one film per year past the age of 100" Sounds
# like he may have made more films after this one as well, but probably didn't
# reach the popularity threshold we put on to create this subset

# Now let's move to the youngest. This needs some research and a few assumptions to be made!
imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(year<=deathYear | is.na(deathYear)) %>%
    filter(DirectorReleaseAge==11)
    

# I am a little suspicious of how an 11 year old is directing Egyptian comedies:
# https://www.imdb.com/title/tt2210509/

# Let's check out his IMDB profile: https://www.imdb.com/name/nm12679691/

# In the 2010s he seemed to be working both as a child actor and also a
# director/casting director! Busy schedule to do all of that on top of school
# and stuff... or maybe there are two Abdelrahman Yassers and IMDB doesn't
# realise this because we are right down in some obscure depths of Egyptian
# cinema here where the active user base is a bit lower to spot this sort of
# thing?

# There is this guy, who is a relatively successful Egyptian actor (born in
# 1999): https://www.amazon.co.uk/prime-video/actor/Abdelrahman-Yasser/ 

# And there is this guy who co-directed a few things and is much less well
# known: https://www.linkedin.com/in/yasser-abdelrahman-1400b373 but started
# working for Egyptian television in 1999 so is surely not the same person.

# Sometimes the data is not correct! The data is created by humans! And humans
# make mistakes all the time Ideally we should go back and correct our data (but
# this will ruin my example for future years!) when we find errors

# In this case lets just set the "director" Abdelrahman Yasser's birthyear to be
# unknown using case_when() inside of mutate(). Since we already have so many
# missing values this one doesn't change much.


imdb %>%
  mutate(birthYear = case_when(director=="Abdelrahman Yasser"~NA,
                                              .default=birthYear)) %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(year<=deathYear | is.na(deathYear)) %>%
      summarise(Youngest=min(DirectorReleaseAge, na.rm=TRUE),  Oldest=max(DirectorReleaseAge, na.rm=TRUE))
    
# So let's take a look again. A 18 year old director sounds a lot more plausible
# to me!

imdb %>%
  mutate(DirectorReleaseAge=year-birthYear) %>%
    filter(year<=deathYear | is.na(deathYear)) %>%
    filter(DirectorReleaseAge==18)
    
# https://en.wikipedia.org/wiki/Samira_Makhmalbaf 
# "At the age of 17, after
# directing two video productions, Makhmalbaf went on to direct her first
# feature film, La Pomme (The Apple)." So she was 17 at time of directing the film and 18 at time of
# release. 
# Which all matches up - we made it!