Load packages

library(knitr) # to put this on my website
library(dplyr) 
library(ggplot2) 
library(here) #to define paths

## here() starts at /Users/sophiadrewry/Documents/School/Archive/Fall 2021/MADA/SophiaPDrewry-MADA-Portfolio

library(lubridate) #to arrange date form of data

Load data

This data can be found on FiveThirtyEight’s github page. I used the adjusted verion Now lets load the data

dataspot1 <- here::here("data", "covid_approval_polls_adjusted.csv")
covid_approval_polls_adjusted <- read.csv(dataspot1)
# This dataset includes the approval and disapproval scores
dataspot2 <- here::here("data", "covid_approval_toplines.csv")
covid_approval_toplines <- read.csv(dataspot2) 
# This data set is for the line

Clean data

Lets get an overview of the data

str(covid_approval_polls_adjusted)

## 'data.frame':    2782 obs. of  19 variables:
##  $ subject            : chr  "Biden" "Biden" "Biden" "Biden" ...
##  $ modeldate          : chr  "9/21/2021" "9/21/2021" "9/21/2021" "9/21/2021" ...
##  $ party              : chr  "D" "D" "D" "D" ...
##  $ startdate          : chr  "1/24/2021" "1/28/2021" "1/29/2021" "1/31/2021" ...
##  $ enddate            : chr  "1/26/2021" "2/1/2021" "2/1/2021" "2/2/2021" ...
##  $ pollster           : chr  "YouGov" "Quinnipiac University" "Morning Consult" "YouGov" ...
##  $ grade              : chr  "B+" "A-" "B" "B+" ...
##  $ samplesize         : num  477 333 808 484 564 ...
##  $ population         : chr  "a" "a" "rv" "a" ...
##  $ weight             : num  0.629 0.632 0.834 0.549 0.888 ...
##  $ influence          : num  0 0 0 0 0 0 0 0 0 0 ...
##  $ multiversions      : chr  "" "" "" "" ...
##  $ tracking           : logi  NA NA NA NA NA NA ...
##  $ approve            : num  84 93 89 88 89.2 ...
##  $ disapprove         : num  3 5 7 7 7.14 8 5 5 3 2 ...
##  $ approve_adjusted   : num  86.9 92 90.3 90.9 89.6 ...
##  $ disapprove_adjusted: num  2.43 5.7 6.4 6.43 6.68 ...
##  $ timestamp          : chr  "20:55:08 21 Sep 2021" "20:55:08 21 Sep 2021" "20:55:08 21 Sep 2021" "20:55:08 21 Sep 2021" ...
##  $ url                : chr  "https://docs.cdn.yougov.com/ld46rgtdlz/econTabReport.pdf" "https://poll.qu.edu/national/release-detail?ReleaseID=3688" "https://assets.morningconsult.com/wp-uploads/2021/02/03074817/2101110_crosstabs_POLITICO_RVs_v1.pdf" "https://docs.cdn.yougov.com/460mactkmh/econTabReport.pdf" ...

str(covid_approval_toplines)

## 'data.frame':    3300 obs. of  6 variables:
##  $ subject            : chr  "Trump" "Biden" "Trump" "Biden" ...
##  $ modeldate          : chr  "9/21/2021" "9/21/2021" "9/21/2021" "9/21/2021" ...
##  $ party              : chr  "all" "all" "R" "R" ...
##  $ approve_estimate   : num  38.9 52 76.8 19.5 35.9 ...
##  $ disapprove_estimate: num  57 42.7 20 76.9 57.5 ...
##  $ timestamp          : chr  "20:55:28 21 Sep 2021" "20:55:22 21 Sep 2021" "20:55:21 21 Sep 2021" "20:55:17 21 Sep 2021" ...

Lets make the data more managable

biden <- covid_approval_polls_adjusted %>% 
     filter(subject %in% c("Biden")) %>%
     filter(party %in% c("all")) %>%
     select(startdate, approve_adjusted, disapprove_adjusted)

Convert character into date format using lubridate package

# Well after an hour of running code that does not work, I figured out that the problem was that the date is not in "date" form
# headbang
# headbang
# headbang

biden$startdate <- mdy(biden$startdate)

Now lets build out plot

# Hex code for colors are 
# Orange = #EF813F
# Purple = #D78CDB

s1 <- ggplot() + 
     geom_point(data = biden, mapping=aes(x=startdate, y=approve_adjusted), alpha = 0.2, color = "#EF813F") + 
     geom_point(data = biden, mapping=aes(x=startdate, y=disapprove_adjusted), alpha = 0.2, color = "#D78CDB") +
     xlab(" ") +
     ylab(" ") + ylim(20, 80)

# No x or y axis titles because there are none in the model graph
print(s1)

Now to do a similar process with the line data

bidenline <- covid_approval_toplines %>% 
     filter(subject %in% c("Biden")) %>%
     filter(party %in% c("all")) %>%
     select(modeldate, approve_estimate, disapprove_estimate)

#Proper date format
bidenline$modeldate <- mdy(bidenline$modeldate) # Will not forget this again!

#Now for the plot
l1 = ggplot() + 
    geom_line(data = bidenline, aes(x = modeldate, y = approve_estimate), color = "#EF813F") +
    geom_line(data = bidenline, aes(x = modeldate, y = disapprove_estimate), color = "#D78CDB") +
    xlab(" ") +
    ylab(" ") + ylim(20, 80)

print(l1)

Ok this is looking good. Now to combine both graphs

BigGraph = ggplot() + 
    geom_point(data = biden, mapping=aes(x=startdate, y=approve_adjusted), alpha = 0.3, color = "#EF813F") + 
    geom_point(data = biden, mapping=aes(x=startdate, y=disapprove_adjusted), alpha = 0.3, color = "#D78CDB") +
    geom_line(data = bidenline, aes(x = modeldate, y = approve_estimate), color = "#EF813F", size = 1.5) +
    geom_line(data = bidenline, aes(x = modeldate, y = disapprove_estimate), color = "#D78CDB", size = 1.5) +
    xlab(" ") +
    ylab(" ") + ylim(20, 80) +
    scale_x_date( date_breaks = "1 month", date_labels = "%m/%d") +
    theme_bw() +
    annotate("text", x = as.Date("2021-01-21"), y = 80,
           label = "Biden sworn \ninto office",
           color = "black",
           size = 2,
           fontface = 1,
           vjust = 1,
           hjust = 0) +
    annotate("text", x = as.Date("2021-09-21"), y = 42, 
           label = "42.7% Disapprove", 
           color = "#D78CDB", 
           size = 4, 
           fontface = 2,
           vjust = 1,
           hjust = 0) +
    annotate("text", x = as.Date("2021-09-21"), y = 51, 
           label = "51.7% Approve", 
           color = "#EF813F", 
           size = 4, 
           fontface = 2,
           vjust = 1,
           hjust = 0)
print(BigGraph)

Lets compare to original graph

Now to add some theme elements

BigGraph2 = ggplot() + 
    geom_point(data = biden, mapping=aes(x=startdate, y=approve_adjusted), alpha = 0.3, color = "#EF813F") + 
    geom_point(data = biden, mapping=aes(x=startdate, y=disapprove_adjusted), alpha = 0.3, color = "#D78CDB") +
    geom_line(data = bidenline, aes(x = modeldate, y = approve_estimate), color = "#EF813F", size = 1.5) +
    geom_line(data = bidenline, aes(x = modeldate, y = disapprove_estimate), color = "#D78CDB", size = 1.5) +
    xlab(" ") +
    ylab(" ") + ylim(20, 80) +
    theme_light (base_size=12, base_family="mono") +
    theme(
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),  
        axis.line.y.left = element_line(color = "#D3D3D3", linetype = 2),
        axis.line.y.right = element_line(color =    "#D3D3D3", linetype = 2),
        axis.line.x = element_line(color =  "#D3D3D3"),
        axis.text = element_text(colour = "#D3D3D3"),
        axis.title = element_text(colour = "#D3D3D3"),
        axis.ticks = element_line(colour = "#D3D3D3"),
        aspect.ratio=2/6) +
    scale_x_date( date_breaks = "1 month", date_labels = "%m/%d") +
    annotate("text", x = as.Date("2021-01-21"), y = 80,
           label = "Biden sworn \ninto office",
           color = "black",
           size = 2,
           fontface = 1,
           vjust = 1,
           hjust = 0) +
    annotate("text", x = as.Date("2021-09-21"), y = 42, 
           label = "42.7% Disapprove", 
           color = "#D78CDB", 
           size = 4, 
           fontface = 2,
           vjust = 1,
           hjust = 0) +
    annotate("text", x = as.Date("2021-09-21"), y = 51, 
           label = "51.7% Approve", 
           color = "#EF813F", 
           size = 4, 
           fontface = 2,
           vjust = 1,
           hjust = 0) 
print(BigGraph2)

Save all my hard work

figure_file = here::here("BigGraph2.png")
ggsave(filename = figure_file, plot=BigGraph2, dpi = 300)

## Saving 7 x 5 in image

Visualization

Load packages

Load data

Clean data

Lets make the data more managable

Convert character into date format using lubridate package

Now lets build out plot

Now to do a similar process with the line data

Ok this is looking good. Now to combine both graphs

Now to add some theme elements

Save all my hard work