DATA607_Tidyverse_Project
Josh Iden, extended by Mahmud Hasan Al Raji
2022-10-26
TIDYVERSE PROJECT
This project uses data from fivethirtyeight.com to demonstrate tidyverse package capabilities.
library(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6 ✔ purrr 0.3.4
## ✔ tibble 3.1.8 ✔ dplyr 1.0.10
## ✔ tidyr 1.2.1 ✔ stringr 1.4.1
## ✔ readr 2.1.3 ✔ forcats 0.5.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()THE DATA
url <- "https://raw.githubusercontent.com/fivethirtyeight/data/master/election-deniers/fivethirtyeight_election_deniers.csv"
df <- read.csv(url)
head(df)## Candidate Incumbent State Office District Stance
## 1 Katie Britt No Alabama Senator N/A Fully denied
## 2 Jerry Carl Yes Alabama Representative 1 Fully denied
## 3 Barry Moore Yes Alabama Representative 2 Fully denied
## 4 Mike Rogers Yes Alabama Representative 3 Fully denied
## 5 Robert Aderholt Yes Alabama Representative 4 Fully denied
## 6 Dale Strong No Alabama Representative 5 Fully denied
## Source
## 1 NBC News
## 2 Congressional roll call, Alabama Political Reporter
## 3 Congressional roll call
## 4 Congressional roll call
## 5 Congressional roll call
## 6 Facebook
## URL
## 1 https://twitter.com/VaughnHillyard/status/1528918324192608257
## 2 https://clerk.house.gov/Votes/202111?Date=01%2F07%2F2021
## 3 https://clerk.house.gov/Votes/202111?Date=01%2F07%2F2021
## 4 https://clerk.house.gov/Votes/202111?Date=01%2F07%2F2021
## 5 https://clerk.house.gov/Votes/202111?Date=01%2F07%2F2021
## 6 https://www.facebook.com/strongforalabama/posts/pfbid0Km4TZaJPUTYuSAnwLNeezfrQSaL3uCNDLhvWFwRkMDpKPByKPTXDFYEdvP2m1rvkl?__cft__[0]=AZVE-LgrXbka-w1K8Z5IBxGpBh61GPUdZxnovjIPky71lJcPQWMquAkzM7W8jVhSl7p6idLzrG9toCil8_CQJUa_1NsiaaLVvWTqhVnrllFsloNysJKgcXmgeC0g3E_M4nJnPgsbVJ4ATcVTKt3uUJsRCZRWe028wDT8sg4hQaIRJsMxEFe46wgxhd-UBLyMZ5OKiWp2sNUyyvnmr92-UpxH&__tn__=%2CO%2CP-R
## Note
## 1
## 2
## 3
## 4
## 5
## 6The dplyr package allows us to use pipes to transform the data:
df <- df |>
mutate(Stance = toupper(Stance))It also let’s us count the number of times a value occurs:
stance_ct <- df |>
count(Stance) |>
arrange(desc(n))
stance_ct## Stance n
## 1 FULLY DENIED 199
## 2 NO COMMENT 104
## 3 ACCEPTED WITH RESERVATIONS 93
## 4 FULLY ACCEPTED 77
## 5 RAISED QUESTIONS 61
## 6 AVOIDED ANSWERING 18The ggplot package allows us to visualize this without the transformation.
ggplot(df) +
geom_histogram(aes(x=Stance), stat="count") +
coord_flip()
The tidyr package lets us pivot data:
by_state <- df |> # using dplyr pipe
group_by(State, Stance) |>
summarise(count = n()) |>
pivot_wider(names_from = Stance, values_from = count)## `summarise()` has grouped output by 'State'. You can override using the
## `.groups` argument.by_state## # A tibble: 50 × 7
## # Groups: State [50]
## State ACCEPTED WITH RESERVATI…¹ FULLY…² RAISE…³ FULLY…⁴ NO CO…⁵ AVOID…⁶
## <chr> <int> <int> <int> <int> <int> <int>
## 1 Alabama 1 8 2 NA NA NA
## 2 Alaska 1 1 2 1 NA NA
## 3 Arizona 1 9 2 NA 1 NA
## 4 Arkansas 1 1 1 3 1 1
## 5 California 6 13 2 9 18 1
## 6 Colorado 1 4 NA 6 1 NA
## 7 Connecticut 2 NA 2 2 3 NA
## 8 Delaware NA NA NA NA 2 NA
## 9 Florida 2 19 5 2 2 1
## 10 Georgia 6 8 2 2 NA NA
## # … with 40 more rows, and abbreviated variable names
## # ¹`ACCEPTED WITH RESERVATIONS`, ²`FULLY DENIED`, ³`RAISED QUESTIONS`,
## # ⁴`FULLY ACCEPTED`, ⁵`NO COMMENT`, ⁶`AVOIDED ANSWERING`Extended part summary
I will see the numbers of Republican nominee for different offices who took different stances on the 2020 election. I will also see the office wise nominee numbers who fully denied the election. Finally, I will find the incumbents by types for different stances and will also graphically represent the incumbent numbers by grouped barplot for each stance category.
Subsetting the df dataframe with only Incumbent,Office, and Stance columns by using select function from dplyr
df1<-df %>% select(Incumbent,Office,Stance)
head(df1)## Incumbent Office Stance
## 1 No Senator FULLY DENIED
## 2 Yes Representative FULLY DENIED
## 3 Yes Representative FULLY DENIED
## 4 Yes Representative FULLY DENIED
## 5 Yes Representative FULLY DENIED
## 6 No Representative FULLY DENIEDFinding nominee numbers for different offices who took different stances on the 2020 election by using group_by and summarize functions of dplyr
df2 <- df1 %>%
group_by(Office) %>%
summarize(count = n())
df2## # A tibble: 6 × 2
## Office count
## <chr> <int>
## 1 Attorney general 30
## 2 Governor 36
## 3 Representative 424
## 4 Secretary of state 26
## 5 Senator 35
## 6 Senator (unexpired term) 1Finding the Republican nominee numbers by offices who fully denied the 2020 election by filtering the required data by using filter function of dplyr
df3<-df1 %>% filter(Stance=="FULLY DENIED") %>%
group_by(Office) %>% summarize(count = n())
df3## # A tibble: 6 × 2
## Office count
## <chr> <int>
## 1 Attorney general 7
## 2 Governor 7
## 3 Representative 170
## 4 Secretary of state 7
## 5 Senator 7
## 6 Senator (unexpired term) 1Finding Incumbent numbers for different stance cateory by using group_by annd summarize functions of dplyr package
df3 <- df1 %>%
group_by(Stance,Incumbent) %>%
summarize(count = n())## `summarise()` has grouped output by 'Stance'. You can override using the
## `.groups` argument.df3## # A tibble: 12 × 3
## # Groups: Stance [6]
## Stance Incumbent count
## <chr> <chr> <int>
## 1 ACCEPTED WITH RESERVATIONS No 42
## 2 ACCEPTED WITH RESERVATIONS Yes 51
## 3 AVOIDED ANSWERING No 15
## 4 AVOIDED ANSWERING Yes 3
## 5 FULLY ACCEPTED No 38
## 6 FULLY ACCEPTED Yes 39
## 7 FULLY DENIED No 80
## 8 FULLY DENIED Yes 119
## 9 NO COMMENT No 98
## 10 NO COMMENT Yes 6
## 11 RAISED QUESTIONS No 51
## 12 RAISED QUESTIONS Yes 10Visualizing two types of incumbent numbers for each category of stance by grouped barplot using different functions of ggplot2 package
ggplot(df3, aes(factor(Stance), count,fill=Incumbent)) +
geom_bar(stat="identity", position = "dodge") + theme(axis.text.x = element_text(angle = 90, size = 10))+
scale_fill_brewer(palette = "Set1")