electionGraphs 0.3.4

Create graphs for displaying the result of a election based on a csv-inputfile.

Election Result Graphs

With ElectionResultGraphs you can create either four different graph types or all graphs as one chart for displaying the result of a election based on a csv-inputfile. As input a csv-file needs to be provided. In the initialization the data will be processed as pandas dataframe and graphs from plotly can be exported as image-files.

pip install electionGraphs

Features

• data from the csv-file will be processed and more data points will be calculated (initialization)
• a .getGraph() function can create different types of graphs
• a bar graph showing the results in relative votes can be displayed
• a bar graph showing a comparision of relative votes to last voting period
• a pie chart shows the seat allocation of the parliament
• a stacked bar chart shows possible coalitions
• .createOnePager() creates all graph options and combines them into one chart

Seat Allocation (Parliament)

• seats of a parliament will be allocated based on relative voting results
• a threshold is implemented, so that a party needs to reach a certain relative voting result in order to be considered in the seat allocation (this feature can be adjusted by parameters)
• a list of parties can be excluded from the seat allocation (e.g. a "other" category)
• for the seat allocation the relative results (excluding parties that are defined by the excludeParties parameters or by the percentageLimit) will be multiplied by parliamentSeats, the result will be floored. Not allocated seats will then be allocated each party ordered by the highest modulus (decimal rest from the division) until all seats are allocated

Coalitions

• coalitions will be calculated based on total seats. The political spectrum will be considered as a threshold
• for every possible coaltion the distance based on the delivered information of columnSpectrum will be calculated. Reaches the overall distance the thresholdPoliticalDistance the coalition will not be included in the final graph

Disclaimer

This tool is loosly based on german elections. Some rulesets may differ from the reality (e.g. not considering first vote, second vote). This is just a hobby project for practicing data processing and data visualization.

Example Graph

One Pager with multiple different graphs

Single chart of bar graph

VotingGraphs

For the initialization the mandatory parameters must be given.

from electionGraphs import electionGraphs

# example for instantiating the class

eg = electionGraphs.ElectionGraphs(
    "exampleData.csv",
    "YEAR",
    "VOTINGS",
    "PARTY_SHORT",
    "PARTY_SPEC",
    "PARTY_COLOR",
)

After initialization different aspects of logic and styling can be configured by setting the optional parameters:

votingGraphs.titleMain = "My other Title"
votingGraphs.colors["background"] = "#000000"
votingGraphs.fontsize["titleMain"] = 58

mandatory parameters/attributes

csvFile = (str) csv file to read in -- For example of the structure of csv see example file

columnYear = (str) name of column in csv in which the year is recorded

columnVotings = (str) name of column in csv in which the absolute number of votings per party is recorded

columnParty = (str) name of column in csv in which party name is recorded

columnSpectrum = (str) name of column in csv in which the political orientation of the party is recorded

columnColor = (str) name of column in csv in which the color code of the party is recorded

optional parameters/attributes

colors = (dict{str:str}) hex-color for different elements of the graphs. The value must be a str of hex-color (e.g. "#F2EAD3") with following keys:

• background
• diagram
• title
• subtitle
• yaxis
• xaxis
• grid
• values
• threshold

excludeParties = (list[str], default = ["Sonstiges", "Sonstige", "Other"]) list of party names which will not be included in the seat allocation

filenameBarCompare = (str, default = "barDifference.png") filename of export graph. Must be a .png file

filenameBarResult = (str, default = "barResult.png") filename of export graph. Must be a .png file

filenameBarCoalitions = (str, default = "barCoalition.png") filename of export graph. Must be a .png file

filenameOnePager = (str, default = "ElectionResults") filename of export graph. Must be a .png file

filenamePieParliament = (str, default = "pieParliament.png") filename of export graph. Must be a .png file

fontfamily = (str, default = "Futura") font family of all graphs

fontsize = (dict{str:int}) font sizes of different elements. Each key represent a text type with following options: title subtitle values yaxis xaxis

outputfolder = (str, default = "output") folder in which all output images will be saved

parliamentSeats = (int, default = 120) number of seats in parliament. This parameter needs be defined at the initialization, otherwise some graphs will not display the seats correctly

percentageLimit = (int, default = 5) threshold at which a party will not be considered when calculting seat distribution

thresholdPolitcalDistance = (int, default = 300) defines a threshold for displaying possible coalitions. When the politcal distance in a mathmatical possible coalition is too high, it can be filtered

titleBarCoalitions = (str, default="Koalitionen") adjusts title of bar graph with possible coalitions

titleBarCompare = (str, default="Veränderung der Wählerstimmen") adjusts title of bar graph with comparision to last voting

titleBarResult = (str, default="Wählerstimmen") adjusts title of bar graph with results

titleMain = (str, default="Wahlergebnisse") adjusts main title of one pager

titlePieParliament = (str, default="Sitzverteilung") adjusts title of parliament graph

seperator = (str, default = ";") type of seperator in csv

subtitleBarCoalitions= (str, default="Anzahl Sitze für mögliche Koalitionen") adjusts a subtitle of corresponding graph

subtitleBarCompare = (str, default="Prozentpunkte im Vergleich zur letzten Wahl") adjusts a subtitle of corresponding graph

subtitleBarResult = (str, default="Anteil der Wählerstimmen in Prozent") adjusts a subtitle of corresponding graph

subtitlePieParliament = (str, default="Anzahl Sitze im Parlament") adjusts a subtitle of corresponding graph

function: getGraph

Creates a png-file of a specific chart type for the provided data

# example for creating a graph
electionGraphs.getGraph(2021, type="BAR_DIFFERENCE")

mandatory parameters/attributes

year = (int) year of data

type = (str, specific options) type of graph:

• "BAR_RESULT" = bar graph with basic results

• "BAR_COMPARE" = bar graph with comparison to last voting year

• "PIE_PARLIAMENT" = pie chart for seat distribution in a parliament

• "BAR_COALITIONS" = stacked bar chart with possible coalitions

optional parameters/attributes

Many of the optional parameters/attributes of the main class VotingGraphs will be applied to the charts (e.g. titles, colors, fonts, etc.)

function: createOnePager

Creates a png-file with all four chart types arranged

# example for creating a one pager
electionGraphs.createOnePager()

optional parameters/attributes

year = (int) year of data, if left empty all possible years will be processed and for each year a file will be created

Many of the optional parameters/attributes of the main class VotingGraphs will be applied to the charts (e.g. titles, colors, fonts, etc.)

Dependencies

• pandas
• plotly
• kaleido
• pillow