When you click the Render button a document will be generated that includes both content and the output of embedded code. You can embed code like this:
library(dplyr)
Attaching package: 'dplyr'The following objects are masked from 'package:stats':
filter, lagThe following objects are masked from 'package:base':
intersect, setdiff, setequal, unionlibrary(tidyr)
library(ggplot2)Warning: package 'ggplot2' was built under R version 4.5.2library(scales)
maineresults <- read.csv("maineresults.csv")
mainedemographics <- read.csv("mainedemographics.csv")You can add options to executable code like this
sum(maineresults$rcv_yes)[1] 378831sum(maineresults$rcv_no)[1] 353856sum(maineresults$mw_yes)[1] 412799sum(maineresults$mw_no)[1] 332946sum(maineresults$bg_yes)[1] 361873sum(maineresults$bg_no)[1] 386869rcv_support <- 378831/(378831+353856)
mj_support <-412799/(412799+332946)
bg_support <-361873/(361873+386869)
rcv_support[1] 0.5170434mj_support[1] 0.5535391bg_support[1] 0.483308For the code above I calculated the average support for all three ballot initiatives.
election_results <- maineresults %>%
summarise(
rcv_yes=sum(rcv_yes,na.rm = TRUE),
rcv_no=sum(rcv_no,na.rm = TRUE),
mw_yes = sum(mw_yes, na.rm = TRUE),
mw_no = sum(mw_no, na.rm = TRUE),
bg_yes = sum(bg_yes, na.rm = TRUE),
bg_no = sum(bg_no, na.rm = TRUE)
) %>%
mutate(
rcv_support=rcv_yes/(rcv_yes+rcv_no),
mw_support=mw_yes/(mw_yes+mw_no),
bg_support=bg_yes/(bg_yes+bg_no)
)
election_results rcv_yes rcv_no mw_yes mw_no bg_yes bg_no rcv_support mw_support bg_support
1 378831 353856 412799 332946 361873 386869 0.5170434 0.5535391 0.483308ballot_results <- election_results %>%
select(rcv_support, mw_support, bg_support) %>%
pivot_longer(
everything(),
names_to = "measure",
values_to = "support"
)
ballot_results$measure <- recode(ballot_results$measure,
rcv_support = "Ranked Choice Voting",
mw_support = "Marijuana Legalization",
bg_support = "Background Checks"
)
ggplot(ballot_results, aes(x = measure, y = support)) +
geom_col(fill = "steelblue") +
geom_text(aes(label = percent(support, accuracy = 1)), vjust = -0.4) +
geom_hline(yintercept = 0.5, linetype = "dashed") +
scale_y_continuous(
limits = c(0,1),
breaks = seq(0,1,0.2),
labels = percent_format()
) +
labs(
title = "Statewide Support for Maine Ballot Measures",
x = "Ballot Measure",
y = "Yes Vote Share"
) +
theme_minimal()
As shown above I created a bar graph to visualize the findings. I also created a line in the bar graph at the fifty percent mark, to better illustrate how much support over the fifty percent mark each initiative got.
maineprojections <- read.csv("maineprojections.csv")
sum(maineprojections$proj_rcv_yes)[1] 421842sum(maineprojections$proj_mw_yes)[1] 450264sum(maineprojections$proj_bg_yes)[1] 390712proj_results <- maineprojections %>%
summarise(
proj_votes = sum(proj_votes, na.rm = TRUE),
proj_mw_yes = sum(proj_mw_yes, na.rm = TRUE),
proj_bg_yes = sum(proj_bg_yes, na.rm = TRUE),
proj_rcv_yes = sum(proj_rcv_yes, na.rm = TRUE)
) %>%
mutate(
rcv_support = proj_rcv_yes / proj_votes,
mw_support = proj_mw_yes / proj_votes,
bg_support = proj_bg_yes / proj_votes
)
actual_results <- election_results %>%
select(rcv_support, mw_support, bg_support) %>%
pivot_longer(everything(), names_to = "measure", values_to = "support") %>%
mutate(type = "Actual")
proj_maine <- proj_results %>%
select(rcv_support, mw_support, bg_support) %>%
pivot_longer(everything(), names_to = "measure", values_to = "support") %>%
mutate(type = "Projected")
comparison <- bind_rows(actual_results, proj_maine) %>%
mutate(measure = recode(measure,
rcv_support = "Ranked Choice Voting",
mw_support = "Marijuana Legalization",
bg_support = "Background Checks"))ggplot(comparison, aes(x = measure, y = support, fill = type)) +
geom_col(position = "dodge") +
geom_text(
aes(label = percent(support, accuracy = 1)),
position = position_dodge(width = 0.9),
vjust = -0.4
) +
geom_hline(yintercept = 0.5, linetype = "dashed") +
scale_y_continuous(
limits = c(0, 1),
breaks = seq(0, 1, 0.2),
labels = percent_format()
) +
labs(
title = "Projected vs Actual Support for Maine Ballot Measures",
x = "Ballot Measure",
y = "Yes Vote Share",
fill = "Series"
) +
theme_minimal()
As shown in the graph and codings above, I joined the maineprojections dataset with the maineresults dataset. This then allowed me to compare the projections vs the actual results of each dataset on one graph.
mainedemographics <- mainedemographics %>%
mutate(
county = case_when(
geoid >= 2300102060 & geoid <= 2300179585 ~ "ANDROSCOGGIN",
geoid >= 2300300800 & geoid <= 2300387215 ~ "AROOSTOOK",
geoid >= 2300502655 & geoid <= 2300587845 ~ "CUMBERLAND",
geoid >= 2300702235 & geoid <= 2300785850 ~ "FRANKLIN",
geoid >= 2300901185 & geoid <= 2300986655 ~ "HANCOCK",
geoid >= 2301100590 & geoid <= 2301186970 ~ "KENNEBEC",
geoid >= 2301301465 & geoid <= 2301380425 ~ "KNOX",
geoid >= 2301501010 & geoid <= 2301587075 ~ "LINCOLN",
geoid >= 2301701325 & geoid <= 2301787355 ~ "OXFORD",
geoid >= 2301901115 & geoid <= 2301987390 ~ "PENOBSCOT",
geoid >= 2302100100 & geoid <= 2302185710 ~ "PISCATAQUIS",
geoid >= 2302301570 & geoid <= 2302387460 ~ "SAGADAHOC",
geoid >= 2302501395 & geoid <= 2302582840 ~ "SOMERSET",
geoid >= 2302703950 & geoid <= 2302786760 ~ "WALDO",
geoid >= 2302900380 & geoid <= 2302985290 ~ "WASHINGTON",
geoid >= 2303100275 & geoid <= 2303187985 ~ "YORK",
TRUE ~ NA_character_
)
)
head(mainedemographics) geoid n_registered share_dem share_rep share_white share_afam
1 2302100100 450 0.207 0.413 0.953 0.002
2 2303100275 1905 0.270 0.316 0.927 0.006
3 2302900380 844 0.262 0.379 0.947 0.004
4 2301100590 1394 0.244 0.370 0.952 0.001
5 2302900660 374 0.257 0.329 0.955 0.011
6 2303100730 2287 0.282 0.272 0.932 0.001
share_female avg_hhincome avg_popdens avg_partyscore avg_collegescore
1 0.504 41.030 81.967 40.137 33.673
2 0.493 72.358 406.412 42.668 43.955
3 0.519 62.874 135.724 37.743 40.843
4 0.479 70.391 153.811 39.703 36.418
5 0.476 44.395 133.396 43.884 41.506
6 0.507 60.793 262.589 48.724 43.263
avg_gunownscore avg_gvpscore avg_churchscore avg_marijuanascore
1 68.418 23.165 23.710 54.763
2 60.220 29.071 23.037 63.063
3 67.649 25.181 28.313 58.456
4 68.423 24.403 24.498 60.077
5 70.698 23.934 24.434 62.467
6 55.856 35.669 22.985 62.200
avg_fiscalprogscore avg_choicescore avg_enviroscore county
1 24.368 42.687 20.035 PISCATAQUIS
2 29.105 55.195 25.557 YORK
3 27.567 48.887 23.190 WASHINGTON
4 26.337 50.523 21.824 KENNEBEC
5 28.598 52.162 23.616 WASHINGTON
6 30.863 56.128 28.127 YORKmaineresults <- maineresults %>%
mutate(
county = case_when(
geoid >= 2300102060 & geoid <= 2300179585 ~ "ANDROSCOGGIN",
geoid >= 2300300800 & geoid <= 2300387215 ~ "AROOSTOOK",
geoid >= 2300502655 & geoid <= 2300587845 ~ "CUMBERLAND",
geoid >= 2300702235 & geoid <= 2300785850 ~ "FRANKLIN",
geoid >= 2300901185 & geoid <= 2300986655 ~ "HANCOCK",
geoid >= 2301100590 & geoid <= 2301186970 ~ "KENNEBEC",
geoid >= 2301301465 & geoid <= 2301380425 ~ "KNOX",
geoid >= 2301501010 & geoid <= 2301587075 ~ "LINCOLN",
geoid >= 2301701325 & geoid <= 2301787355 ~ "OXFORD",
geoid >= 2301901115 & geoid <= 2301987390 ~ "PENOBSCOT",
geoid >= 2302100100 & geoid <= 2302185710 ~ "PISCATAQUIS",
geoid >= 2302301570 & geoid <= 2302387460 ~ "SAGADAHOC",
geoid >= 2302501395 & geoid <= 2302582840 ~ "SOMERSET",
geoid >= 2302703950 & geoid <= 2302786760 ~ "WALDO",
geoid >= 2302900380 & geoid <= 2302985290 ~ "WASHINGTON",
geoid >= 2303100275 & geoid <= 2303187985 ~ "YORK",
TRUE ~ NA_character_
)
)As shown above, I created a county variable for both the mainedemographics dataset and the maineresults dataset. I did this because I wanted to compare the results of all three ballot iniatives and demographics within each county.
income_county <- mainedemographics %>%
group_by(county) %>%
summarise(
mean_income = mean(avg_hhincome, na.rm = TRUE)
)
marijuana_county <- maineresults %>%
group_by(county) %>%
summarise(
mw_yes = sum(mw_yes, na.rm = TRUE),
mw_no = sum(mw_no, na.rm = TRUE)
) %>%
mutate(
mw_support = mw_yes / (mw_yes + mw_no)
)
county_results <- income_county %>%
left_join(marijuana_county, by = "county")In this code above I calculated the average income per each county and then caculated the support for the marijuana iniative in each county. Afterwards I combined both datasets.
ggplot(county_results,
aes(x = mean_income, y = mw_support)) +
geom_point(size = 3) +
geom_smooth(method = "lm", se = FALSE) +
scale_y_continuous(labels = percent_format()) +
labs(
title = "Average County Income vs Marijuana Support in Maine",
x = "Average Household Income",
y = "Marijuana Yes Vote Share"
) +
theme_minimal()`geom_smooth()` using formula = 'y ~ x'
ggplot(county_results,
aes(x = mean_income, y = mw_support)) +
geom_point(size = 3) +
geom_text(aes(label = county), vjust = -0.6) +
geom_smooth(method = "lm", se = FALSE) +
scale_y_continuous(labels = percent_format()) +
labs(
title = "Average County Income vs Marijuana Support in Maine",
x = "Average Household Income",
y = "Marijuana Yes Vote Share"
) +
theme_minimal()`geom_smooth()` using formula = 'y ~ x'
In both graphs shown above I added a linear regression line to show the relationship between the two variables.
ggplot(county_results,
aes(x = reorder(county, mw_support), y = mw_support)) +
geom_col(fill = "steelblue") +
coord_flip() +
scale_y_continuous(labels = percent_format()) +
labs(
title = "Marijuana Support by County",
x = "County",
y = "Yes Vote Share"
) +
theme_minimal()
ggplot(county_results,
aes(x = reorder(county, mean_income), y = mean_income)) +
geom_col(fill = "steelblue") +
geom_text(aes(label = dollar(mean_income)), hjust = -0.1) +
coord_flip() +
labs(
title = "Average Household Income by County",
x = "County",
y = "Average Household Income"
) +
theme_minimal()
gun_county <- mainedemographics %>%
group_by(county) %>%
summarise(
mean_gun = mean(avg_gunownscore, na.rm = TRUE)
)
bg_county <- maineresults %>%
group_by(county) %>%
summarise(
bg_yes = sum(bg_yes, na.rm = TRUE),
bg_no = sum(bg_no, na.rm = TRUE)
) %>%
mutate(
bg_support = bg_yes / (bg_yes + bg_no)
)
bg_analysis <- gun_county %>%
left_join(bg_county, by = "county")ggplot(bg_analysis,
aes(x = mean_gun/100, y = bg_support)) +
geom_point(size = 3, color = "black") +
geom_text(
aes(label = county),
size = 3,
nudge_y = 0.01
) +
geom_smooth(method = "lm", se = FALSE, color = "blue") +
scale_x_continuous(
labels = percent_format(),
limits = c(0.45, 0.67)
) +
scale_y_continuous(
labels = percent_format(),
limits = c(0.25, 0.70)
) +
labs(
title = "Gun Ownership vs Support for Background Checks in Maine Counties",
x = "Estimated Gun Ownership Rate by County",
y = "Background Check Yes Vote Share"
) +
theme_minimal()`geom_smooth()` using formula = 'y ~ x'
college_county <- mainedemographics %>%
group_by(county) %>%
summarise(
mean_college = mean(avg_collegescore, na.rm = TRUE)
)
bg_county <- maineresults %>%
group_by(county) %>%
summarise(
bg_yes = sum(bg_yes, na.rm = TRUE),
bg_no = sum(bg_no, na.rm = TRUE)
) %>%
mutate(
bg_support = bg_yes / (bg_yes + bg_no)
)
bg_analysis <- college_county %>%
left_join(bg_county, by = "county")ggplot(bg_analysis, aes(x = mean_college/100, y = bg_support)) +
geom_point(size = 3) +
geom_text(aes(label = county), vjust = -0.6, size = 3) +
geom_smooth(method = "lm", se = FALSE) +
scale_x_continuous(labels = percent_format()) +
scale_y_continuous(labels = percent_format(), limits = c(0,1)) +
labs(
title = "College Attainment vs Support for Background Checks (by County)",
x = "Estimated College Attainment Rate (County)",
y = "Background Check Support"
) +
theme_minimal()`geom_smooth()` using formula = 'y ~ x'