babynames

### Quick illustration of dplyr / ggplot2

library(dplyr)
library(ggplot2)
library(magrittr)

# Install babynames package
install.packages("babynames")

# load library
library(babynames)


## take a look at the data

str(babynames)

head(babynames)

tail(babynames)

?babynames

#  Think of some questions you'd like me to answer using this data ???

### e.g. which name was the most popular ever - and in which year?
babynames %>% arrange(desc(n))

babynames %>% arrange(desc(prop)) 

babynames %>% arrange(desc(prop)) %>% filter(sex=="F")

babynames %>% arrange(desc(prop)) %>% filter(sex=="F")

babynames %>% arrange(desc(prop)) %>% filter(sex=="F" & name!="Mary")


# e.g. How has the use of James changed over time?

james <- babynames %>%
              filter(name=="James") 

head(james)

# line graph
ggplot(james, aes(year, n)) +
    geom_line(aes(color=sex), lwd=1) +
    scale_color_manual(values = c("firebrick1", "dodgerblue")) +
    theme_bw()


#can do it all in one

babynames %>% 
  filter(name=="Chris") %$%
  ggplot(., aes(year, n)) +
    geom_line(aes(color=sex), lwd=1) +
    scale_color_manual(values = c("firebrick1", "dodgerblue")) +
    theme_bw()




## let's compare three names...

threenames <- babynames %>%
               filter(sex=="F") %>%
               filter(name=="Jennifer" | name=="Sarah" | name=="Mary" ) 

head(threenames)

ggplot(threenames, aes(year, n)) +
  geom_line(aes(group=name, color=name), lwd=1) +
  scale_color_manual(values = c("firebrick1", "dodgerblue", "darkorange")) +
  theme_bw()



# another way of doing the filtering above.

mynames <- c("Kim", "Khloe", "Kendall", "Kourtney", "Kylie")

kards <- babynames %>%
  filter(sex=="F") %>%
  filter((name %in% mynames ) == T)

head(kards)

ggplot(kards, aes(year, n)) +
  geom_line(aes(group=name, color=name), lwd=1) +
  scale_color_manual(values = c("firebrick1", "dodgerblue", "darkorange", "purple", "green1")) +
  theme_bw()




## e.g. 2. Names that don't get used anymore...

pre <-
  babynames %>% 
  filter(year<1945) %>%
  group_by(name) %>%
  summarize(total = sum(n)) %>%
  rename(npre = total)


post <-
  babynames %>% 
  filter(year>=1945) %>%
  group_by(name) %>%
  summarize(total = sum(n)) %>%
  rename(npost = total)

head(pre)
head(post)

full_join(pre, post, by = "name")  #notice that some rows have "NA" - means they're missing (i.e. zeros)

ournames <- full_join(pre, post, by = "name")  #notice that some rows have "NA" - means they're missing (i.e. zeros)

ournames %>%
  filter(is.na(npost)==T) %>%
  arrange(desc(npre))


## let's look at distribution of each of these in one graph....

#extact top 10 vanished names as a character vector
oldnames <- ournames %>%
  filter(is.na(npost)==T) %>%
  arrange(desc(npre)) %>%
  head(10) %>% 
  .$name

oldnames 



oldnames1 <- babynames %>%
  filter((name %in% oldnames ) == T)

head(oldnames1)

ggplot(oldnames1, aes(year, n)) +
  geom_line(aes(group=name, color=name), lwd=1) +
  theme_bw()



### names beginning with "Adelaid"

x<-babynames[(grepl("Adelaid", babynames$name) == T),]
unique(x$name)

adelaids <- 
  babynames %>%
  filter(grepl("Adelaid", name) == T) %>%
  filter(sex=="F") %>% 
  group_by(year) %>%
  summarize(total=sum(n))

adelaids

ggplot(adelaids, aes(year, total)) +
  geom_line(color="black", lwd=1) +
  theme_bw()