| Type: | Package |
| Title: | Bayesian "Now-Cast" Estimation of Event Probabilities in Multi-Party Democracies |
| Version: | 0.6.27 |
| Date: | 2026-04-12 |
| Maintainer: | Andreas Bender <bender.at.R@gmail.com> |
| Description: | An implementation of a Bayesian framework for the opinion poll based estimation of event probabilities in multi-party electoral systems (Bender and Bauer (2018) <doi:10.21105/joss.00606>). |
| Depends: | R (≥ 3.2.1) |
| Imports: | checkmate, gtools, rvest, xml2, rlang, magrittr, lubridate, stringr, tidyr (≥ 1.0.0), purrr (> 0.2.2), dplyr (> 0.5.0), ggplot2, tibble (≥ 3.0.0) |
| Suggests: | testthat, covr, |
| Encoding: | UTF-8 |
| License: | MIT + file LICENSE |
| URL: | https://adibender.github.io/coalitions/ |
| BugReports: | https://github.com/adibender/coalitions/issues |
| RoxygenNote: | 7.3.2 |
| LazyData: | true |
| NeedsCompilation: | no |
| Packaged: | 2026-04-12 18:35:18 UTC; abender |
| Author: | Andreas Bender 
[aut, cre],
Alexander Bauer
[aut],
Rebekka Schade [ctb] |
| Repository: | CRAN |
| Date/Publication: | 2026-05-08 15:10:15 UTC |
Pipe operator
Description
See magrittr::%>% for details.
Usage
lhs %>% rhs
Value
The result of applying rhs to lhs.
Creates basic survey table from votes in percent
Description
This functions takes votes in percent (per party) obtained from a survey, and returns a table containing votes (in percent) and party names. Conducts sanity checks along the way, such as checking that percentages add up to 1.
Usage
as_survey(
percent,
samplesize,
parties = c("cdu", "spd", "gruene", "fdp", "linke", "piraten", "afd", "fw", "sonstige"),
epsilon = 1e-05
)
Arguments
percent |
Votes in percent each party received in the survey
of interest. Can be set to |
samplesize |
Number of respondents in survey. |
parties |
Vector of same length and in the same order as |
epsilon |
The parameter |
Value
A data.frame containing input and absolute number of votes
in survey per party.
See Also
Examples
forsa <- as_survey(
percent = c(0.41, 0.24, 0.13, 0.04, 0.08, 0.03, 0.03, 0.04),
samplesize = 2508,
parties = c("cdu/csu", "spd", "gruene", "fdp", "linke", "piraten", "afd", "others"))
forsa
Calculate coalition probability from majority table
Description
Given a table with simulations in the rows and coalitions in the columns, this function returns the coalition probabilities for a specified coalition, by default excluding superior coalitions first
Usage
calculate_prob(majority_df, coalition, exclude_superior = TRUE, ...)
Arguments
majority_df |
A data frame containing logical values indicating if the coalitions (columns) have a majority (rows). |
coalition |
The coalition of interest for which superior coalitions
will be obtained by |
exclude_superior |
Logical. If |
... |
Further arguments passed to |
Value
A data frame with one numeric column giving the coalition probability (percentage of simulations in which the coalition obtained a majority, after optionally excluding superior coalitions).
Examples
test_df <- data.frame(
cdu = c(rep(FALSE, 9), TRUE),
cdu_fdp = c(rep(FALSE, 8), TRUE, TRUE),
cdu_fdp_greens = c(TRUE, TRUE, rep(FALSE, 6), TRUE, TRUE))
calculate_prob(test_df, "cdu_fdp_greens") # exclude_superior defaults to TRUE
calculate_prob(test_df, "cdu_fdp_greens", exclude_superior=FALSE)
Calculate coalition probabilities for multiple coalitions
Description
Given a table with simulations in the rows and coalitions in the columns, this function returns the coalition probabilities for a specified coalition, by default excluding superior coalitions first
Usage
calculate_probs(majority_df, coalitions, exclude_superior = TRUE, ...)
Arguments
majority_df |
A data frame containing logical values indicating if the coalitions (columns) have a majority (rows). |
coalitions |
A list of coalitions for which coalition probabilities should
be calculated. Each list entry must be a vector of party names. Those names
need to correspond to the names in |
exclude_superior |
Logical. If |
... |
Further arguments passed to |
Value
A data frame with columns coalition (character) and
probability (numeric, 0–100), one row per coalition.
See Also
Examples
test_df <- data.frame(
cdu = c(rep(FALSE, 9), TRUE),
cdu_fdp = c(rep(FALSE, 8), TRUE, TRUE),
cdu_fdp_greens = c(TRUE, TRUE, rep(FALSE, 6), TRUE, TRUE))
calculate_probs(test_df, list("cdu", "cdu_fdp", "cdu_fdp_greens"))
calculate_probs(test_df, list("cdu", "cdu_fdp", "cdu_fdp_greens"), exclude_superior=FALSE)
Transform surveys in long format
Description
Given a data frame containing multiple surveys (one row per survey), transforms the data into long format with one row per party.
Usage
collapse_parties(
surveys,
parties = c("cdu", "spd", "greens", "fdp", "left", "pirates", "fw", "afd", "bsw",
"others")
)
Arguments
surveys |
A data frame with one survey per row. |
parties |
A character vector containing names of parties to collapse. |
Value
Data frame in long format
Examples
emnid <- scrape_wahlrecht()
emnid.long <- collapse_parties(emnid)
Seat Distribution by D'Hondt
Description
Calculates number of seats for the respective parties according to the method of d'Hondt.
Usage
dHondt(votes, parties, n_seats = 183)
Arguments
votes |
Number of votes per party. |
parties |
Names of parties (must be same length as votes). |
n_seats |
Number of seats in parliament. Defaults to 183 (seats in Austrian parliament). |
Value
A named integer vector of seat counts, one entry per party,
in the same order as parties. The vector has a logical attribute
ties: TRUE if two or more parties had equal claim to the last
seat (i.e. the result is not uniquely determined and was resolved randomly),
FALSE otherwise. When ties = TRUE, re-running with a different
random seed may produce a different but equally valid seat distribution.
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1) %>% tidyr::unnest("survey")
# calculate the seat distribution based on D'Hondt for a parliament with 300 seats
dHondt(surveys$votes, surveys$party, n_seats = 300)
Draw random numbers from posterior distribution
Description
Draw random numbers from posterior distribution
Usage
draw_from_posterior(
survey,
nsim = 10000,
seed = as.numeric(now()),
prior = NULL,
correction = NULL
)
Arguments
survey |
survey object as returned by |
nsim |
number of simulations |
seed |
sets seed |
prior |
optional prior information. Defaults to 1/2 (Jeffrey's prior). |
correction |
A positive number. If not |
Value
data.frame containing random draws from Dirichlet distribution
which can be interpreted as election results.
See Also
Calculate the effective sample size
Description
This is the work horse function that calculates the effective sample size. Should usually not be called by the user directly.
Usage
effective_samplesize(size, share, corr = 0.5, weights = NULL)
Arguments
size |
A vector of sample sizes from different surveys (from different pollsters) for one party. |
share |
The relative share of votes for party of interest ([0-1]) |
corr |
Assumed correlation between surveys (of different pollsters). Defaults to 0.5. |
weights |
Additional weights for individual surveys. |
Value
A single numeric value: the effective sample size of the pooled sample accounting for the correlation between pollsters.
Extract numerics from string or character
Description
Removes all characters that are not in [0-9].
Usage
extract_num(x, decimal = TRUE)
Arguments
x |
A character vector. |
decimal |
Logical flag, indicating if x has a decimal separator |
Value
A numeric vector with non-numeric characters removed.
Remove rows from table for which superior coalitions are possible
Description
Given a table with simulations in the rows and coalitions in the columns, this function returns the coalition probabilities for a specified coalition, by default excluding superior coalitions first
Usage
filter_superior(majority_df, coalition, ...)
Arguments
majority_df |
A data frame containing logical values indicating if the coalitions (columns) have a majority (rows). |
coalition |
The coalition of interest for which superior coalitions
will be obtained by |
... |
Further arguments passed to |
Value
A data frame with the same structure as majority_df but with
rows removed where any superior coalition also has a majority.
See Also
Examples
test_df <- data.frame(
cdu = c(rep(FALSE, 9), TRUE),
cdu_fdp = c(rep(FALSE, 8), TRUE, TRUE),
cdu_fdp_greens = c(TRUE, TRUE, rep(FALSE, 6), TRUE, TRUE))
calculate_prob(test_df, "cdu_fdp_greens") # exclude_superior defaults to TRUE
calculate_prob(test_df, "cdu_fdp_greens", exclude_superior=FALSE)
Extract surveys from institutes within a specified time-window
Description
Extract surveys from institutes within a specified time-window
Usage
get_eligible(
surveys,
pollsters,
last_date = Sys.Date(),
period = 14,
period_extended = NA
)
Arguments
surveys |
A |
pollsters |
Character vector of pollsters that should be considered for pooling. |
last_date |
Only surveys in the time-window from |
period |
See |
period_extended |
Optional. If specified, all surveys in the time-window
from |
Value
A tibble with one row per pollster containing the most recent survey within the specified time window, filtered and down-weighted as appropriate.
Get probabilities to enter the parliament.
Description
Get probabilities to enter the parliament.
Usage
get_entryprobability(dirichlet.draws, hurdle = 0.05)
Arguments
dirichlet.draws |
Matrix or data frame containing draws from
the posterior (see |
hurdle |
The percentage threshold which has to be reached by a party to enter the parliament. Any party called "ssw" will be exempt from the hurdle. |
Value
Vector of (named) entry probabilities.
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1)
# use 100 simulations for a fast runtime
surveys <- surveys %>% mutate(draws = purrr::map(survey, draw_from_posterior, nsim = 100),
entryProbs = purrr::map(draws, get_entryprobability))
surveys$entryProbs
Extract "meta" information from survey data base
Description
Extract "meta" information from survey data base
Usage
get_meta(surveys_df)
Arguments
surveys_df |
A data frame containing surveys from different survey
institutes as returned by |
Value
A tibble with columns pollster, date, start,
end, and respondents (one row per survey).
Total number of survey participants from surveys eligible for pooling.
Description
Total number of survey participants from surveys eligible for pooling.
Usage
get_n(eligible_df)
Arguments
eligible_df |
A data frame containing surveys that should be used for
pooling as returned by |
Value
A single numeric value: the total number of respondents across all eligible surveys (one survey per pollster, after down-weighting).
Extract effective sample size for pooled sample
Description
Given a specified time window (defaults to current day - 14 days). calculate the effective sample size of the pooled sample over multiple pollsters.
Usage
get_pooled(
surveys,
last_date = Sys.Date(),
pollsters = c("allensbach", "emnid", "forsa", "fgw", "gms", "infratest", "dimap",
"infratestdimap", "insa"),
period = 14,
period_extended = NA,
corr = 0.5,
weights = NULL
)
Arguments
surveys |
A |
last_date |
Only surveys in the time-window from |
pollsters |
Character vector of pollsters that should be considered for pooling. |
period |
See |
period_extended |
Optional. If specified, all surveys in the time-window
from |
corr |
Assumed correlation between surveys (of different pollsters). Defaults to 0.5. |
weights |
Additional weights for individual surveys. |
Value
A tibble with one row per party containing columns party,
from, to, Neff (effective sample size), and
pollsters (comma-separated names of pollsters used).
Wrapper for calculation of coalition probabilities from survey
Description
Given a table with simulations in the rows and coalitions in the columns, this function returns the coalition probabilities for a specified coalition, by default excluding superior coalitions first
Usage
get_probabilities(
x,
coalitions = list(c("cdu"), c("cdu", "fdp"), c("cdu", "fdp", "greens"), c("spd"),
c("spd", "left"), c("spd", "left", "greens")),
nsim = 1e+05,
distrib.fun = sls,
seats_majority = 300L,
seed = as.numeric(now()),
correction = NULL
)
Arguments
x |
A table containing one row per survey and survey information in
long format in a separate column named |
coalitions |
A list of coalitions for which coalition probabilities should
be calculated. Each list entry must be a vector of party names. Those names
need to correspond to the names in |
nsim |
number of simulations |
distrib.fun |
Function to calculate seat distribution. Defaults to
|
seats_majority |
The number of seats needed to obtain majority. |
seed |
sets seed |
correction |
A positive number. If not |
Value
A tibble with the same rows as x (one per survey) and an
additional list-column probabilities containing a data frame of
coalition names and their probabilities (0–100) for each survey.
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1)
# calculate probabilities for two coalitions
probs <- get_probabilities(surveys,
coalitions = list(c("cdu", "fdp"),
c("spd", "left", "greens")),
nsim = 100) # ensure fast runtime with only 100 simulations
probs %>% tidyr::unnest("probabilities")
Calculate seat distribution from draws from posterior
Description
Calculate seat distribution from draws from posterior
Usage
get_seats(
dirichlet.draws,
survey,
distrib.fun = sls,
samplesize = NULL,
hurdle = 0.05,
others = "others",
...
)
Arguments
dirichlet.draws |
Matrix containing random draws from posterior. |
survey |
The actual survey results on which |
distrib.fun |
Function to calculate seat distribution. Defaults to
|
samplesize |
Number of individuals participating in the |
hurdle |
The percentage threshold which has to be reached by a party to enter the parliament. Any party called "ssw" will be exempt from the hurdle. |
others |
A string indicating the name under which parties not listed explicitly are subsumed. |
... |
Further arguments passed to |
Value
A data frame containing seat distributions for each simulation in
dirichlet.draws
See Also
draw_from_posterior, sls,
dHondt
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1)
# simulate 100 seat distributions
surveys <- surveys %>% mutate(draws = purrr::map(survey, draw_from_posterior, nsim = 100),
seats = purrr::map2(draws, survey, get_seats))
surveys$seats
Extract superior coalitions from coalition string or vector
Description
Extract superior coalitions from coalition string or vector
Usage
get_superior(string, pattern = "_", collapse = "_")
Arguments
string |
A character. |
pattern |
Pattern to look for (regular expression). |
collapse |
string that will be used to concatenate multiple elements
obtained by splitting |
Value
A character vector of all proper subsets (superior coalitions) of
the parties in string.
See Also
stringr str_split
Scrape surveys from all pollsters
Description
Given a specific date, extract the survey from this date or the last one before this date.
Usage
get_surveys(country = "DE")
get_surveys_by()
get_surveys_rp()
get_surveys_nds()
get_surveys_saxony()
get_surveys_brb()
get_surveys_thuringen()
get_latest(surveys = NULL, max_date = Sys.Date())
Arguments
country |
Choose country from which surveys should be scraped.
Currently |
surveys |
If provided, latest survey will be obtained from this object,
otherwise calls |
max_date |
Specifies the date, relative to which latest survey will
be searched for. Defaults to |
Value
Nested tibble. When fully unnested, the dataset contains the following columns:
- pollster
Character name of the polling institute.
- date
Publication date of the poll.
- start, end
Start and end date of the field period, i.e. the dates during which the poll was conducted.
- respondents
Number of respondents in the poll.
- party
Character name of an individual party.
- percent
Percentage of respondents that chose the party. Given in percentage points, i.e.
38%is given as38.- votes
Number of respondents that chose the party.
Examples
library(coalitions)
get_surveys()
library(coalitions)
### Scrape the newest poll for the German federal election
# Possibility 1: Calling get_latest without arguments scrapes surveys from the web
# Possibility 2: Use get_latest() on an already scraped dataset
surveys <- get_latest(surveys_sample)
Plot voter shares observed in one survey
Description
Bar chart of the raw voter shares observed in one survey. Additionally to plotting positive voter shares, the function can be used to plot party-specific differences (e.g. between a survey and the election result), including negative numbers.
Usage
gg_survey(data, colors = NULL, labels = NULL, annotate_bars = TRUE, hurdle = 5)
Arguments
data |
Scraped dataset containing one row per party in the column
|
colors |
Named vector containing party colors. If |
labels |
Named vector containing party labels. If |
annotate_bars |
If |
hurdle |
Hurdle for single parties to get into the parliament, e.g. '5'
for '5%'. If set to NULL no horizontal line is plotted.
The horizontal line can be suppressed using |
Value
A ggplot object displaying voter shares as a bar chart.
Examples
library(tidyr)
library(dplyr)
library(coalitions)
survey <- surveys_sample$surveys[[1]]$survey[[1]]
gg_survey(survey)
Seat Distribution by Hare/Niemeyer
Description
Calculates number of seats for the respective parties that have received more
than hurdle percent of votes (according to the method of Hare/Niemeyer)
Usage
hare_niemeyer(votes, parties, n_seats = 183)
Arguments
votes |
Number of votes per party. |
parties |
Names of parties (must be same length as votes). |
n_seats |
Number of seats in parliament. Defaults to 183 (seats in Austrian parliament). |
Value
A data.frame containing parties above the hurdle and the respective
seats/percentages after redistribution via Hare/Niemeyer
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1) %>% tidyr::unnest("survey")
# calculate the seat distribution based on Hare/Niemeyer for a parliament with 300 seats
hare_niemeyer(surveys$votes, surveys$party, n_seats = 300)
Does a coalition have a majority
Description
Does a coalition have a majority
Usage
has_majority(seats_tab, coalition, seats_majority = 300L)
Arguments
seats_tab |
A table containing information on how many seats each party obtained. |
coalition |
The coalition of interest for which superior coalitions
will be obtained by |
seats_majority |
The number of seats needed to obtain majority. |
Value
A data frame with one logical column majority indicating
whether the coalition obtained a majority in each simulation.
Do coalitions have a majority
Description
Do coalitions have a majority
Usage
have_majority(
seats_tab,
coalitions = list(c("cdu"), c("cdu", "fdp"), c("cdu", "fdp", "greens"), c("spd"),
c("spd", "left"), c("spd", "left", "greens")),
seats_majority = 300L,
collapse = "_"
)
Arguments
seats_tab |
A data frame containing number of seats obtained by a party.
Must have columns |
coalitions |
A list of coalitions for which coalition probabilities should
be calculated. Each list entry must be a vector of party names. Those names
need to correspond to the names in |
seats_majority |
The number of seats needed to obtain majority. |
collapse |
Character string passed to |
Value
A data frame with one column per coalition. Each column is logical indicating whether the coalition obtained a majority in each simulation row.
Examples
library(coalitions)
library(dplyr)
library(purrr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1)
# check for majorities of two coalitions
coals <- list(c("cdu", "fdp"),
c("spd", "left", "greens"))
# only use 100 simulations for a fast runtime
surveys <- surveys %>% mutate(draws = map(survey, draw_from_posterior, nsim = 100),
seats = map2(draws, survey, get_seats),
majorities = map(seats, have_majority, coalitions = coals))
surveys$majorities
Colors for German parties
Description
A vector of colors associated with German parties.
Usage
party_colors_de
Format
A named character vector. Names indicate parties. Values contain color strings for the respective parties
Labels for German parties
Description
A vector of labels associated with German parties.
Usage
party_labels_de
Format
A named character vector. Names indicate parties. Values contain party names suitable for plot labels.
Transform list of coalitions to vector by combining party names
Description
Transform list of coalitions to vector by combining party names
Usage
paste_coalitions(coalitions, collapse = "_")
Arguments
coalitions |
A list of coalitions for which coalition probabilities should
be calculated. Each list entry must be a vector of party names. Those names
need to correspond to the names in |
collapse |
Character string passed to |
Value
A character vector of coalition names formed by concatenating party
names with collapse.
Obtain pooled survey during specified period
Description
Per default, pools surveys starting from current date and going 14 days back. For each pollster within the defined time-frame, only the most recent survey is used.
Usage
pool_surveys(
surveys,
last_date = Sys.Date(),
pollsters = c("allensbach", "emnid", "forsa", "fgw", "gms", "infratest", "dimap",
"infratestdimap", "insa"),
period = 14,
period_extended = NA,
corr = 0.5,
weights = NULL
)
Arguments
surveys |
A |
last_date |
Only surveys in the time-window from |
pollsters |
Character vector of pollsters that should be considered for pooling. |
period |
See |
period_extended |
Optional. If specified, all surveys in the time-window
from |
corr |
Assumed correlation between surveys (of different pollsters). Defaults to 0.5. |
weights |
Additional weights for individual surveys. |
Value
A data frame with one row per party containing columns
pollster (set to "pooled"), date, start,
end, respondents (effective sample size), party,
percent, and votes.
Examples
library(coalitions)
library(dplyr)
latest <- get_latest(surveys_sample)
pool_surveys(surveys_sample, last_date=as.Date("2017-09-02"))
Replace/prettify matching words/terms in one vector by another
Description
The function searches for x values, that occur in current
and replaces them with entries in new. Useful for quick
renaming/translation of survey column names and by using internal object
.trans_df
Usage
prettify_strings(
x,
current = .trans_df$english,
new = .trans_df$english_pretty
)
prettify_de(x)
prettify_en(x)
Arguments
x |
A character vector (or factor) that should be renamed. |
current |
A vector of characters (possibly subset of |
new |
A vector of characters that will replace entries in |
Value
A character vector (or factor, if input was factor) with matched
entries replaced by the corresponding values in new.
Examples
library(coalitions)
library(dplyr)
# look at sample German federal election polls
surveys <- surveys_sample %>% tidyr::unnest("surveys") %>% group_by(pollster) %>% slice(1)
# prettify the polling agency names
prettify_strings(surveys$pollster)
prettify_en(surveys$pollster)
prettify_de(surveys$pollster)
Calculate percentage of votes/seats after excluding parties with
votes < hurdle
Description
Calculate percentage of votes/seats after excluding parties with
votes < hurdle
Usage
redistribute(survey, hurdle = 0.05, others = "others", epsilon = 1e-05)
Arguments
survey |
The actual survey results on which |
hurdle |
The percentage threshold which has to be reached by a party to enter the parliament. Any party called "ssw" will be exempt from the hurdle. |
others |
A string indicating the name under which parties not listed explicitly are subsumed. |
epsilon |
Percentages should add up to 1. If they do not, within accuracy
of |
Value
A data frame with the same structure as survey but with
parties below the hurdle removed and vote percentages renormalized.
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1)
# redistribute the shares of 'others' parties and parties with a share of under 5\%
surveys <- surveys %>% mutate(survey_redist = purrr::map(survey, redistribute))
surveys$survey # results before redistribution
surveys$survey_redist # results after redistribution
Sanitize column names
Description
Sanitize column names
Usage
sanitize_colnames(df)
Arguments
df |
A data frame with party names with special characters that need to be sanitized. |
Value
The input data frame with column names converted to lowercase ASCII and party name abbreviations standardized.
Sanitize character vector
Description
Substitute all German "Umlaute"
Usage
sanitize_strings(x)
Arguments
x |
A character vector. |
Value
A character vector with German umlauts replaced by their ASCII equivalents (e.g. "oe" for "รถ").
Scrape surveys for German general election
Description
Scrapes survey tables and performs sanitation to output tidy data
Usage
scrape_wahlrecht(
address = "https://www.wahlrecht.de/umfragen/emnid.htm",
parties = c("CDU", "SPD", "GRUENE", "FDP", "LINKE", "PIRATEN", "AFD", "BSW",
"SONSTIGE")
)
scrape_by(
address = "https://www.wahlrecht.de/umfragen/landtage/bayern.htm",
parties = c("CSU", "SPD", "GRUENE", "FDP", "LINKE", "PIRATEN", "FW", "AFD", "SONSTIGE")
)
scrape_rp(
address = "https://www.wahlrecht.de/umfragen/landtage/rheinland-pfalz.htm",
parties = c("CDU", "SPD", "GRUENE", "FDP", "LINKE", "AFD", "FW", "SONSTIGE"),
ind_row_remove = -c(1:3)
)
scrape_ltw(
address = "https://www.wahlrecht.de/umfragen/landtage/niedersachsen.htm",
parties = c("CDU", "SPD", "GRUENE", "FDP", "LINKE", "PIRATEN", "FW", "AFD", "BSW",
"SONSTIGE"),
ind_row_remove = -c(1:2)
)
Arguments
address |
http-address from which tables should be scraped. |
parties |
A character vector containing names of parties to collapse. |
ind_row_remove |
Negative vector of rows that will be skipped at the beginning. |
Value
A tibble with one row per survey date and columns for date, respondents, and one column per party containing the percentage of votes.
Examples
library(coalitions)
library(dplyr)
scrape_wahlrecht() %>% slice(1:5)
# Niedersachsen
scrape_ltw() %>% slice(1:5)
# Hessen
scrape_ltw("https://www.wahlrecht.de/umfragen/landtage/hessen.htm", ind_row_remove=-c(1)) %>%
slice(1:5)
Seat Distribution by Sainte-Lague/Schepers
Description
Calculates number of seats for the respective parties that have received more than 5% of votes (according to the method of Sainte-Lague/Schepers, see https://www.wahlrecht.de/verfahren/rangmasszahlen.html).
Usage
sls(votes, parties, n_seats = 598L)
Arguments
votes |
A numeric vector giving the redistributes votes |
parties |
A character vector indicating the names of parties with
respective |
n_seats |
The total number of seats that can be assigned to the different parties. |
Value
A named integer vector of seat counts, one entry per party,
in the same order as parties. The vector has a logical attribute
ties: TRUE if two or more parties had equal claim to the last
seat (i.e. the result is not uniquely determined and was resolved randomly),
FALSE otherwise. When ties = TRUE, re-running with a different
random seed may produce a different but equally valid seat distribution.
See Also
Examples
library(coalitions)
library(dplyr)
# get the latest survey for a sample of German federal election polls
surveys <- get_latest(surveys_sample) %>% ungroup() %>% slice(1) %>% tidyr::unnest("survey")
# calculate the seat distribution based on Sainte-Lague/Schepers for a parliament with 300 seats
sls(surveys$votes, surveys$party, n_seats = 300)
Sample of selected surveys
Description
A data set with surveys from seven different pollsters, three surveys per pollster. Surveys report support for different parties in the running for the German Bundestag prior to the 2017 election.
Usage
surveys_sample
Format
A nested data frame with 7 rows and 2 columns:
- institute
name of the pollster
- surveys
a list of data frames, each containing one survey
Source
Try call of read_html that throws an error if the url cannot be resolved
Description
Try call of read_html that throws an error if the url cannot be resolved
Usage
try_readHTML(url)
Arguments
url |
http-address that should be scraped. |
Value
An xml_document object as returned by xml2::read_html.