# Problem 4, Chapter 6

# Answers may vary.

# Problem 6, Chapter 6:

library(mosaicData)
head(HELPfull)

#(a)
HELPfullsmall <- HELPfull %>%
 filter(ID %in% 1:3) %>%
 select(ID, TIME, DRUGRISK, SEXRISK) 

#(c)
HELPfullwide <- HELPfullsmall %>%
 pivot_wider(names_from = TIME, values_from = c(DRUGRISK,SEXRISK)) %>%
 select(DRUGRISK_0,DRUGRISK_6,SEXRISK_0,SEXRISK_6)

#(a)
HELPfullall <- HELPfull %>%
 select(ID, TIME, DRUGRISK, SEXRISK) %>%
 pivot_wider(names_from = TIME, values_from = c(DRUGRISK,SEXRISK)) %>%
 select(DRUGRISK_0,DRUGRISK_6,SEXRISK_0,SEXRISK_6)

cor(HELPfullall$DRUGRISK_0,HELPfullall$DRUGRISK_6,use="complete.obs")

# [1] 0.5991146

cor(HELPfullall$SEXRISK_0,HELPfullall$SEXRISK_6,use="complete.obs")

# [1] 0.5076615

# Problem 7, Chapter 6:

library(readr)
ds1 <- read_csv(file="https://people.stat.sc.edu/hitchcock/prob7chap6.csv")

# Verify this works:
Treat <- filter(ds1, group == "T")
Control <- filter(ds1, group == "C")
all <- mutate(Treat, diff = Treat$vals - Control$vals)
all

# Better approach:
ds2 <- ds1 %>%
 pivot_wider(names_from=group, values_from=vals) %>%
 mutate(diff = T-C)

# Problem 10, Chapter 6:

library(Lahman)

CubsTeams <- Teams %>%
select(teamID, yearID, HR, HRA) %>%
filter(teamID=='CHN')

CubsTeamslong <- CubsTeams %>%
 pivot_longer(-c(teamID,yearID), names_to = "type", values_to = "HRs")

ggplot(
  data = CubsTeamslong, 
  aes(x = yearID, y = HRs, color = type) ) +
  geom_line()

# Problem 1, Chapter 7:

library(mosaicData)

## avoids warning...
HELPrct %>%
  summarize(across(where(is.numeric), \(x) mean(x, na.rm = TRUE)) )

# Problem 4, Chapter 7 Exercises

# The count_seasons function takes a teamID (INCLUDING quotation marks) as its input,
# and returns an integer number of seasons played for that team as its output.
count_seasons <- function(x){
teamsum <- Teams %>%  
   group_by(teamID) %>%
   summarise(
    num_seasons=n()
    ) %>%
   filter(teamID==x)
ans <- teamsum$num_seasons
return(ans)
}


library(Lahman)
bk_teams <- c("BR1", "BR2", "BR3", "BR4", "BRO", "BRP", "BRF")

map_int(bk_teams, count_seasons)

# Nicer display of results:

Brooklyn_count <- map_int(bk_teams, count_seasons)
print(data.frame(bk_teams, Brooklyn_count))