question7.R

library(tidyverse)

elections = tribble(
  ~year,~state_po,~party_Detailed,~candidatevotes,~totalvotes,
  1976,"AL","DEMOCRAT",500,1000,
  1976,"AL","REPUBLICAN",450,1000,
  1976,"AL","x",30,1000,
  1976,"AL","x",10,1000,
  1976,"AL","x",7,1000,
  1976,"AL","x",2,1000,
  1976,"AL","x",1,1000,
  1976,"AK","REPUBLICAN",70,100,
  1976,"AK","DEMOCRAT",20,100,
  1976,"AK","x",10,100,
)

# Question 7: 
# Consider again the elections data set shown on the previous page. 
# We would like to compute the vote share of each of the two major parties 
# in each year and state, e.g. the votes for that party in that state in that year, 
# divided by the total number of votes for both parties in that state. 

# Which of the following lines of code achieves this goal? 

# Option B, i.e. the answer initially marked as 'incorrect'
filter(elections, party_Detailed %in% c("DEMOCRAT", "REPUBLICAN")) |>   
  group_by(year, state_po) |> 
  summarize(share=candidatevotes / sum(candidatevotes))

# Option D, i.e. the answer initially marked as 'incorrect'
filter(elections, party_Detailed %in% c("DEMOCRAT", "REPUBLICAN")) |>   
  group_by(year, state_po) |> 
  mutate(share=candidatevotes / sum(candidatevotes))

results_b.txt

# A tibble: 4 × 3
# Groups:   year, state_po [2]
   year state_po share
  <dbl> <chr>    <dbl>
1  1976 AK       0.778
2  1976 AK       0.222
3  1976 AL       0.526
4  1976 AL       0.474

results_d.txt

# A tibble: 4 × 6
# Groups:   year, state_po [2]
   year state_po party_Detailed candidatevotes totalvotes share
  <dbl> <chr>    <chr>                   <dbl>      <dbl> <dbl>
1  1976 AL       DEMOCRAT                  500       1000 0.526
2  1976 AL       REPUBLICAN                450       1000 0.474
3  1976 AK       REPUBLICAN                 70        100 0.778
4  1976 AK       DEMOCRAT                   20        100 0.222