Polls for the presidential general election — presidential

Scraped from FiveThirtyEight's website of most recent polls on July 28 2020.

presidential_polls

Format

A tibble with 5312 rows and 37 columns

question_id: numeric
poll_id: numeric
cycle: numeric
state: character
pollster_id: numeric
pollster: character
sponsor_ids: numeric
sponsors: character
display_name: character
pollster_rating_id: numeric
pollster_rating_name: character
fte_grade: character
sample_size: numeric
population: character
population_full: character
methodology: character
office_type: character
seat_number: numeric
seat_name: logical
start_date: Date
end_date: Date
election_date: Date
sponsor_candidate: character
internal: logical
partisan: character
tracking: logical
nationwide_batch: logical
ranked_choice_reallocated: logical
created_at: character
notes: character
url: character
stage: character
race_id: numeric
answer: character
candidate_name: character
candidate_party: character
pct: numeric

Source

https://projects.fivethirtyeight.com/polls-page/president_polls.csv

Examples

library(ggplot2)
library(dplyr)

data(presidential_polls)
presidential_polls %>% 
  filter(candidate_party %in% c("DEM", "REP")) %>% 
  filter(!is.na(state)) %>%
  ggplot(aes(x = end_date, y = pct)) + 
    geom_point(aes(colour = candidate_party)) +
    facet_wrap(~state) +
    scale_colour_party()

# national polls
presidential_polls %>% 
  filter(candidate_party %in% c("DEM", "REP")) %>% 
  filter(answer %in% c("Biden", "Trump")) %>% 
  filter(is.na(state)) %>%
  ggplot(aes(x = end_date, y = pct, 
             colour = candidate_party)) + 
    geom_point(alpha = 0.3) +
    geom_smooth(se = FALSE, method="gam") +
    scale_colour_party() +
    theme(legend.position="bottom")
#> `geom_smooth()` using formula 'y ~ s(x, bs = "cs")'