In RStelzle/rstfuns: A Collection Of Functions I Wrote For Various Projects Of Mine

knitr::opts_chunk$set(dpi=300,fig.width = 11)

This is my personal R-Package containing various short functions I wrote for various projects, mostly concerning topics of political science. My main reasons for writing this package are a) to have written a R package once and b) to have an easy, transparent and reproducible way of reusing code between project. The idea for this small package was born after I read this great blog-post by Hilary Parker. This other post of hers was a great help.

Installation:

Installation of rstfuns requires devtools. If devtools is installed rstfuns can be installed using:

devtools::install_github("https://github.com/RStelzle/rstfuns", dependencies = T)
library(rstfuns)

Available functions:

Roll-Call-Data:

Roll-Call-Data is one of the workhorse-data-sources of the sociology of legaslatures, as it is -depending on the legislative body of interest- relatively accessible. One topic of interest to be tackled with Roll-Call-Data might be cohesion or discipline (for a distinction of the two see @ozbudunPartyCohesionWestern1970) of parliamentary groups. For questions of this kind rstfuns contains two functions calc_attina_index() and calc_rice_index(). Both indices can be used to measure cohesion within a set of legislators for a given vote.

That being said, Roll-Call-Data should not be analysed without a word of caution: As roll-call-votes are in most cases not staged randomly, any inferences taken from a sample of roll-call-votes on the population of all votes are built on shaky grounds. For a great overview see @carrubbaRecordUnrecordedLegislative2006.

calc_attina_index()

With M being the count of votes given for the option with the most votes and N being the total count of votes given, including votes to abstain.

calc_attina_index() is used to calculate "Attina's index of agreement" [see @attinaVotingBehaviourEuropean1990c, pp. 564f.]. Attina's index of agreement can be used for votes with the option to abstain from the vote, like in the European Paraliament. calc_attina_index() takes a vector (max. length() = 3) of the amount/share of votes for "yes", "no" and "abstain". When the vector is shorter than length() = 3 it is assumed that no legislators voted for the missing option(s).

Attina's index of agreement results values from -33.333 to 100. A value of 100 represents a unanimous vote. Positive values result, if more than half of the votes were casted for one and the same option. -33.33 means, that equal shares of legislatures voted for "yes", "no" and "abstain".

calc_attina_index() takes the optional argument rescale (default = FALSE). If set to TRUE, calc_attina_index()'s results get rescaled to a scale of 0-1 (for exaple used in @hixPowerPartiesCohesion2005). A value of 0.25 now represents that 50% of the values were casted for one of the options.

calc_rice_index()

With P being the count of "Yes"-, C the count of "No" and N the total count of votes.

calc_rice_index() is used to calculate the "Rice Index" as defined in @riceBehaviorLegislativeGroups1925 [p. 62f.]. It can be used for votes without an option to abstain. Therefore calc_rice_index() only takes a vector of max. length() = 2. The Rice Index varies between 1 (all votes where casted for the same option) and 0 (votes where evenly split between "yes" and "no").

Utility stuff

plot_turnout_parlgov()

rstfuns::plot_turnout_parlgov(c("Germany", "Poland", "France", "United Kingdom"), german = F)

plot_turnout_parlgov() can be used for quick and dirty plotting of election turnout in countries. Data from the ParlGov Database [@doringParliamentsGovernmentsDatabase2019] is used. The function takes one optional argument. If set to german = T countrynames will be translated to German and "and" will be replaced by "und".

enumerate()

enumerate() can be used to print all entries of a vector to one large string enumerating all entries and adding "and" before the last entry. The optional argument german = T changes "and" to "und". The optional argument oxford = T adds an additional (oxford-) comma before the "and".

rstfuns::enumerate(c("Germany", "Italy", "Poland"))

References

RStelzle/rstfuns documentation built on Oct. 3, 2020, 4:06 p.m.