manystates is a package within the many universe of packages useful
for working with states and state-like entities. It contains an ensemble
of datasets on states and state-like entities in the world, including
information on states’ beginning and, where applicable, end dates, and
geographical and other characteristics. Please also check out {manydata}
for more information about the other packages and tools to handle data
from the many universe of packages.
In addition, the package includes several unique functions to help
you work with state data, including code_states() to
identify states within texts, and generate_states() to
create fictional state names. The package is geared towards global
governance research, but can also be used by anyone interested in state
actors across time.
It is easy to install and start analysing global governance data in R. Simply install the core package, manydata, as follows, and then you can discover, install and update various ‘many packages’ from the console.
# install.packages("manydata") # install the core package
manydata::call_packages() # this prints a list of the publicly available data packages currently available
# manydata::call_packages("manystates") # this downloads and installs the named packageOnce you have installed the package, you can access the primary
datacube in the package, states, as follows. You can see
that currently the datacube combines three datasets: Gleditsch and
Ward’s (1999) state list, the International System(s) Dataset (ISD) by
Griffiths and Butcher (2013), and the GGO state list developed by the
PANARCHIC team for this package.
manydata::describe_datacube(manystates::states)## [1] "The `manystates::states` datacube is a list containing 3 datasets: ISD, GW, and GGO"manystates::states## $ISD
## # A tibble: 499 × 36
## stateID StateName Begin End StateNameAlt Latitude Longitude StartType
## <chr> <chr> <mda> <mda> <chr> <chr> <chr> <dbl>
## 1 DEU Germany 0006… 1945… Prussia 52 31 1… 13 24 18… 3
## 2 KOR Korea 0676… 1905… Koryo, Chos… 37 34 0… 126 58 0… 4
## 3 JPN Japan 0710… 1945… <NA> 35 41 2… 139 41 3… 1
## 4 CHI Chien Khouang,… 0800… 1831… <NA> 19 19 2… 103 21 3… 2
## 5 KHM Cambodia 0802… 1863… Khmer Kingd… 11 34 1… 104 55 1… 1
## 6 FRA France 0843… 1942… <NA> 48 51 2… 2 21 8 E 3
## 7 DNK Denmark 0950… 1940… <NA> 55 40 3… 12 34 6 E 1
## 8 GBR United Kingdom 1066… 2016… <NA> 51 30 2… 0 7 39 W 1
## 9 TON Tonga 1100… 1900… Kingdom Of … 21 8 0 S 175 12 0… 4
## 10 OWO Owo 1150… 1893… <NA> 7 11 0 N 5 35 0 E 1
## # ℹ 489 more rows
## # ℹ 28 more variables: EndType <dbl>, cowID <chr>, cowNR <dbl>,
## # ISD_Category <dbl>, Region <dbl>, Start_Am <dbl>, EStart_Am <dbl>,
## # Declare <dbl>, DecDate <chr>, Population <chr>, PopDate <dbl>, PopAm <dbl>,
## # PopulationHigh <dbl>, PopulationLow <dbl>, StartType_Am <dbl>,
## # StartSettle <dbl>, End_Am <dbl>, EndType_Am <dbl>, EndSettle <dbl>,
## # Sovereignty_Am <dbl>, EuroDip <dbl>, Borders <dbl>, Borders_Am <dbl>, …
##
## $GW
## # A tibble: 216 × 7
## stateID StateName Begin End StateNameAlt cowID cowNR
## <chr> <chr> <mdate> <mdate> <chr> <chr> <chr>
## 1 AFG Afghanistan ..1816-01-01 1888-12-30 <NA> AFG 700
## 2 AUH Austria-Hungary ..1816-01-01 1918-11-13 <NA> AUH 300
## 3 BAD Baden ..1816-01-01 1871-01-17 <NA> BAD 267
## 4 BAV Bavaria ..1816-01-01 1871-01-17 <NA> BAV 245
## 5 CHE Switzerland ..1816-01-01 2017-12-31.. <NA> SWZ 225
## 6 CHN China ..1816-01-01 2017-12-31.. <NA> CHN 710
## 7 DEU Germany ..1816-01-01 1945-05-07 Prussia GMY 255
## 8 DNK Denmark ..1816-01-01 2017-12-31.. <NA> DEN 390
## 9 DZA Algeria ..1816-01-01 1830-07-05 <NA> ALG 615
## 10 ESP Spain ..1816-01-01 2017-12-31.. <NA> SPN 230
## # ℹ 206 more rows
##
## $GGO
## # A tibble: 409 × 13
## stateID StateName Capital Begin End Latitude Longitude Region StateNameAlt
## <chr> <chr> <chr> <mda> <mda> <dbl> <dbl> <chr> <chr>
## 1 ITA Italy Rome -075… 9999… 41.9 12.5 Weste… <NA>
## 2 SMR San Marino San Ma… 0301… 9999… 43.9 12.4 South… <NA>
## 3 BGR Bulgaria Sofia 0681… 1018… 42.7 23.3 South… <NA>
## 4 ABK Abkhazia Anakop… 0778… 1008… 43.0 41.0 Weste… Kingdom of …
## 5 DNK Denmark Copenh… 0793… 9999… 55.7 12.6 North… <NA>
## 6 FRA France Paris 0843… 9999… 48.9 2.35 Weste… <NA>
## 7 HRV Croatia Zagreb 0879… 1102… 45.8 16.0 South… <NA>
## 8 CHM Chamba Chamba 0920… 1846… 32.6 76.1 South… <NA>
## 9 GBR United Ki… London 0927… 9999… 51.5 -0.128 Weste… <NA>
## 10 DRV Annam Hue 0939… 1883… 16.5 108. South… <NA>
## # ℹ 399 more rows
## # ℹ 4 more variables: CapitalAlt <chr>, Coder <chr>, Comments <chr>,
## # Source <chr>Working with ensembles of related data has many advantages for robust analysis. For more information on how to work with ensembles of data, please see the manydata package website.
While some ‘many packages’ can and do include novel data, much of what they offer involves standing on the shoulders of giants. ‘many packages’ endeavour to be as transparent as possible about where data comes from, how it has been coded and/or relabelled, and who has done the work. As such, we make it easy to cite the datasets you use by listing the official references using the function above, as well as the package providers for their work assembling the data by using the function below.
For the Gleditsch and Ward (GW) dataset:
Gleditsch, Kristian S., and Michael D. Ward. 1999. “A revised list of independent states since the congress of Vienna.” International Interactions 25: 393-413.
For the International System(s) Dataset (ISD):
Butcher, Charles R., and Ryan D. Griffiths. 2020. “States and their international relations since 1816: introducing version 2 of the International System(s) Dataset (ISD)”. International Interactions, 46(2), 291–308.
In addition to the datasets, the package also includes two main functions that help you work with state data.
code_states() is useful for identifying states or
state-like entities within text data. This is a non-trivial task, as
state names can vary widely in their spelling, abbreviation, and
context, and there are many potential ambiguities (e.g., “Korea” could
refer to North or South Korea). The function uses a comprehensive
library of contemporary and historical state names and their variants to
identify states in a given text. It returns a data frame with the
identified states and their corresponding ISO 3166 alpha-3 codes or
standardised names.
It can handle both single strings and vectors of strings. By default
it will identify the first state mentioned in each text, but it is
possible to search for multiple states by setting the
max_count argument to a value greater than 1.
manystates::code_states(c("The event took place in the United Kingdom.",
"Switzerland and New Zealand agreed that...",
"One of the oldest princely states is Chamba."),
max_count = 2)## [1] "GBR" "{CHE,NZL}" "CHM"The function can currently discern 500 different states or state-like
entities. This is likely to increase in future releases, but is already
2.2 times more than the next best package, {states}, for
instance.
Additionally, there is a seemingly whimsical function that generates a number of random state names. This function is inspired by observed state names, but is not intended to reflect any actual states. It is primarily designed for use in class exercises when a number of novel state names are required. It can also be used to generate fictional state names for creative writing or games.
manystates::generate_states(3)It is based on a Markov model of syllable patterns found in real state names. Your mileage with the results of this function may vary, but it should get more realistic as we add more state names to the underlying library, and in the meantime it might even be instructive for classroom exercises.
Feedback on either of these functions is most welcome.
Development on this package has been funded by the Swiss National Science Foundation (SNSF) Grant Number 188976: “Power and Networks and the Rate of Change in Institutional Complexes” (PANARCHIC). If you find the package useful, please cite the package.
If you have already developed a dataset salient to this package, please reach out by flagging this as an issue for us, or by forking, further developing the package yourself, and opening a pull request so that your data can be made available at the next opportunity.