peacesciencer: Tools and Data for Quantitative Peace Science

peacesciencer is an R package including various functions and data sets to allow easier analyses in the field of quantitative peace science. The goal is to provide an R package that reasonably approximates what made EUGene so attractive to scholars working in the field of quantitative peace science in the early 2000s. EUGene shined because it encouraged replications of conflict models while having the user also generate data from scratch. Likewise, this R package will offer tools to approximate what EUGene did within the R environment (i.e. not requiring Windows for installation).

Installation

You can install this on CRAN, as follows:

install.packages("peacesciencer")

You can install the development version of this package through the devtools package. The development version of the package invariably has more goodies, but may or may not be at various levels of stress-testing.

devtools::install_github("svmiller/peacesciencer")

What’s Included in {peacesciencer}

The package is already well developed and its functionality continues to expand. The current development version has the following functions.

The current development version also includes the following data.

How to Use {peacesciencer}

{peacesciencer} has a user’s guide that is worth reading. The workflow is going to look something like this. This is a “tidy”-friendly approach to a data-generating process in quantitative peace science.

First, start with one of two processes to create either dyad-year or state-year data. The dyad-year data are created with the create_dyadyears() function. It has a few optional parameters with hidden defaults. The user can specify what kind of state system (system) data they want to use—either Correlates of War ("cow") or Gleditsch-Ward ("gw"), whether they want to extend the data to the most recently concluded calendar year (mry) (i.e. Correlates of War state system membership data are current as of Dec. 31, 2016 and the script can extend that to the end of the most recently concluded calendar year), and whether the user wants directed or non-directed dyad-year data (directed).

The create_stateyears() works much the same way, though “directed” and “non-directed” make no sense in the state-year context. Both functions default to Correlates of War state system membership data to the most recently concluded calendar year.

Thereafter, the user can specify what additional variables they want added to these dyad-year or state-year data. Do note: the additional functions lean primarily on Correlates of War state code identifiers. Indeed, the bulk of the quantitative peace science data ecosystem is built around the Correlates of War project. The variables the user wants are added in a “pipe” in a process like this. Do note that the user may want to break up the data-generating process into a few manageable “chunks” (e.g. first generating dyad-year data and saving to an object, adding to it piece by piece).

All told, the process will look something like this. Assume you want to create some data for something analogous to a “dangerous dyads” design for all non-directed dyad-years. Here’s how you’d do it in {peacesciencer}, which is going to be lifted from the source R scripts for the user’s guide. The first part of this code chunk will lean on core {peacesciencer} functionality whereas the other stuff is some post-processing and, as a bonus, some modeling.

# library(tidyverse) # load this first for most/all things
# library(peacesciencer) # the package of interest
# library(stevemisc) # a dependency, but also used for standardizing variables for better interpretation
library(tictoc)

tic()
create_dyadyears(directed = FALSE, mry = FALSE) %>%
  filter_prd() %>%
  add_gml_mids(keep = NULL) %>%
  add_peace_years() %>%
  add_nmc() %>%
  add_democracy() %>%
  add_cow_alliance() %>%
  add_sdp_gdp() -> Data


Data %>%
  mutate(landcontig = ifelse(conttype == 1, 1, 0)) %>%
  mutate(cowmajdyad = ifelse(cowmaj1 == 1 | cowmaj2 == 1, 1, 0)) %>%
  # Create estimate of militarization as milper/tpop
  # Then make a weak-link
  mutate(milit1 = milper1/tpop1,
         milit2 = milper2/tpop2,
         minmilit = ifelse(milit1 > milit2,
                           milit2, milit1)) %>%
  # create CINC proportion (lower over higher)
  mutate(cincprop = ifelse(cinc1 > cinc2,
                           cinc2/cinc1, cinc1/cinc2)) %>%
  # create weak-link specification using Quick UDS data
  mutate(mindemest = ifelse(xm_qudsest1 > xm_qudsest2,
                            xm_qudsest2, xm_qudsest1)) %>%
  # Create "weak-link" measure of jointly advanced economies
  mutate(minwbgdppc = ifelse(wbgdppc2011est1 > wbgdppc2011est2,
                             wbgdppc2011est2, wbgdppc2011est1)) -> Data

# r2sd() is in {stevemisc}, a {peacesciencer} dependency.
# This is just for a more readable regression output.
Data %>%
  mutate_at(vars("cincprop", "mindemest", "minwbgdppc", "minmilit"),
            ~r2sd(.)) -> Data

broom::tidy(modDD <- glm(gmlmidonset ~ landcontig + cincprop + cowmajdyad + cow_defense +
               mindemest + minwbgdppc + minmilit +
               gmlmidspell + I(gmlmidspell^2) + I(gmlmidspell^3), data= Data,
             family=binomial(link="logit")))
#> # A tibble: 11 x 5
#>    term               estimate   std.error statistic   p.value
#>    <chr>                 <dbl>       <dbl>     <dbl>     <dbl>
#>  1 (Intercept)      -3.04      0.0634         -47.9  0        
#>  2 landcontig        1.05      0.0568          18.5  1.26e- 76
#>  3 cincprop          0.446     0.0363          12.3  9.89e- 35
#>  4 cowmajdyad        0.141     0.0575           2.45 1.41e-  2
#>  5 cow_defense      -0.0993    0.0576          -1.72 8.50e-  2
#>  6 mindemest        -0.492     0.0524          -9.38 6.55e- 21
#>  7 minwbgdppc        0.283     0.0509           5.56 2.77e-  8
#>  8 minmilit          0.261     0.0231          11.3  1.33e- 29
#>  9 gmlmidspell      -0.147     0.00507        -29.1  2.51e-186
#> 10 I(gmlmidspell^2)  0.00249   0.000135        18.4  2.05e- 75
#> 11 I(gmlmidspell^3) -0.0000116 0.000000895    -13.0  1.22e- 38
toc()
#> 11.511 sec elapsed

Here is how you might do a standard civil conflict analysis using Gleditsch-Ward states and UCDP conflict data.

tic()
create_stateyears(system = 'gw') %>%
  filter(year %in% c(1946:2019)) %>%
  add_ucdp_acd(type=c("intrastate"), only_wars = FALSE) %>%
  add_peace_years() %>%
  add_democracy() %>%
  add_creg_fractionalization() %>%
  add_sdp_gdp() %>%
  add_rugged_terrain() -> Data

create_stateyears(system = 'gw') %>%
  filter(year %in% c(1946:2019)) %>%
  add_ucdp_acd(type=c("intrastate"), only_wars = TRUE) %>%
  add_peace_years() %>%
  rename_at(vars(ucdpongoing:ucdpspell), ~paste0("war_", .)) %>%
  left_join(Data, .) -> Data

Data %>%
  arrange(gwcode, year) %>%
  group_by(gwcode) %>%
  mutate_at(vars("xm_qudsest", "wbgdppc2011est",
                 "wbpopest"), list(l1 = ~lag(., 1))) %>%
  rename_at(vars(contains("_l1")),
            ~paste("l1", gsub("_l1", "", .), sep = "_") ) -> Data

modCW <- list()
broom::tidy(modCW$"All UCDP Conflicts" <- glm(ucdponset ~ l1_wbgdppc2011est + l1_wbpopest  +
                    l1_xm_qudsest + I(l1_xm_qudsest^2) +
                    newlmtnest + ethfrac + relfrac +
                    ucdpspell + I(ucdpspell^2) + I(ucdpspell^3), data=subset(Data),
                  family = binomial(link="logit")))
#> # A tibble: 11 x 5
#>    term                 estimate std.error statistic  p.value
#>    <chr>                   <dbl>     <dbl>     <dbl>    <dbl>
#>  1 (Intercept)        -5.10      1.35         -3.77  0.000161
#>  2 l1_wbgdppc2011est  -0.285     0.110        -2.59  0.00952 
#>  3 l1_wbpopest         0.229     0.0672        3.41  0.000645
#>  4 l1_xm_qudsest       0.257     0.181         1.43  0.154   
#>  5 I(l1_xm_qudsest^2) -0.726     0.211        -3.44  0.000574
#>  6 newlmtnest          0.0549    0.0666        0.824 0.410   
#>  7 ethfrac             0.442     0.358         1.23  0.217   
#>  8 relfrac            -0.389     0.402        -0.969 0.333   
#>  9 ucdpspell          -0.0738    0.0393       -1.88  0.0601  
#> 10 I(ucdpspell^2)      0.00443   0.00205       2.16  0.0304  
#> 11 I(ucdpspell^3)     -0.0000602 0.0000280    -2.15  0.0316

broom::tidy(modCW$"Wars Only"  <- glm(war_ucdponset ~ l1_wbgdppc2011est + l1_wbpopest  +
                    l1_xm_qudsest + I(l1_xm_qudsest^2) +
                    newlmtnest + ethfrac + relfrac +
                    war_ucdpspell + I(war_ucdpspell^2) + I(war_ucdpspell^3), data=subset(Data),
                  family = binomial(link="logit")))
#> # A tibble: 11 x 5
#>    term                 estimate std.error statistic p.value
#>    <chr>                   <dbl>     <dbl>     <dbl>   <dbl>
#>  1 (Intercept)        -6.59      2.08         -3.16  0.00157
#>  2 l1_wbgdppc2011est  -0.343     0.172        -1.99  0.0463 
#>  3 l1_wbpopest         0.272     0.106         2.56  0.0105 
#>  4 l1_xm_qudsest      -0.0846    0.270        -0.313 0.754  
#>  5 I(l1_xm_qudsest^2) -0.761     0.352        -2.16  0.0307 
#>  6 newlmtnest          0.342     0.112         3.05  0.00226
#>  7 ethfrac             0.333     0.554         0.601 0.548  
#>  8 relfrac            -0.281     0.593        -0.474 0.635  
#>  9 war_ucdpspell      -0.111     0.0562       -1.98  0.0478 
#> 10 I(war_ucdpspell^2)  0.00466   0.00252       1.85  0.0643 
#> 11 I(war_ucdpspell^3) -0.0000499 0.0000302    -1.65  0.0982

toc()
#> 3.892 sec elapsed

Citing What You Do in {peacesciencer}

You can (and should) cite what you do in {peacesciencer}. The package includes a data frame of a BibTeX file (ps_bib) and a function for finding and returning BibTeX entries that you can include in your projects. This is the ps_cite() function. The ps_cite() function takes a string and does a partial match for relevant keywords (as KEYWORDS) associated with entries in the ps_bib file. For example, you can (and should) cite the package itself.

ps_cite("peacesciencer")
#> @Manual{peacesciencer-package,
#>   Author = {Steven V. Miller},
#>   Title = {peacesciencer}: A User's Guide for Quantitative Peace Science in R},
#>   Year = {2021},
#>   Keywords = {peacesciencer, add_capital_distance(), add_ccode_to_gw(), add_gwcode_to_cow(), capitals},
#>   Url = {http://svmiller.com/peacesciencer/}
#> }

You can see what are the relevant citations to consider using for the data returned by add_democracy()

ps_cite("add_democracy()")
#> @Unpublished{coppedgeetal2020vdem,
#>   Author = {Michael Coppedge and John Gerring and Carl Henrik Knutsen and Staffan I. Lindberg and Jan Teorell and David Altman and Michael Bernhard and M. Steven Fish and Adam Glynn and Allen Hicken and Anna Luhrmann and Kyle L. Marquardt and Kelly McMann and Pamela Paxton and Daniel Pemstein and Brigitte Seim and Rachel Sigman and Svend-Erik Skaaning and Jeffrey Staton and Agnes Cornell and Lisa Gastaldi and Haakon Gjerl{\o}w and Valeriya Mechkova and Johannes von R{\"o}mer and Aksel Sundtr{\"o}m and Eitan Tzelgov and Luca Uberti and Yi-ting Wang and Tore Wig and Daniel Ziblatt},
#>   Note = {Varieties of Democracy ({V}-{D}em) Project},
#>   Title = {V-Dem Codebook v10},
#>   Year = {2020},
#>   Keywords = {add_democracy(), v-dem, varieties of democracy}
#> }
#> 
#> 
#> @Unpublished{marshalletal2017p,
#>   Author = {Monty G. Marshall and Ted Robert Gurr and Keith Jaggers},
#>   Note = {University of Maryland, Center for International Development and Conflict Management},
#>   Title = {Polity {IV} Project: Political Regime Characteristics and Transitions, 1800-2016},
#>   Year = {2017},
#>   Keywords = {add_democracy(), polity}
#> }
#> 
#> 
#> @Unpublished{marquez2016qme,
#>   Author = {Xavier Marquez},
#>   Note = {Available at SSRN: http://ssrn.com/abstract=2753830},
#>   Title = {A Quick Method for Extending the {U}nified {D}emocracy {S}cores},
#>   Year = {2016},
#>   Keywords = {add_democracy(), UDS, Unified Democracy Scores},
#>   Url = {http://dx.doi.org/10.2139/ssrn.2753830}
#> }
#> 
#> 
#> @Article{pemsteinetal2010dc,
#>   Author = {Pemstein, Daniel and Stephen A. Meserve and James Melton},
#>   Journal = {Political Analysis},
#>   Number = {4},
#>   Pages = {426--449},
#>   Title = {Democratic Compromise: A Latent Variable Analysis of Ten Measures of Regime Type},
#>   Volume = {18},
#>   Year = {2010},
#>   Keywords = {add_democracy(), UDS, Unified Democracy Scores},
#>   Owner = {steve},
#>   Timestamp = {2011.01.30}
#> }

You can also return partial matches to see what citations are associated with, say, alliance data in this package.

ps_cite("alliance")
#> @Article{leedsetal2002atop,
#>   Author = {Bretty Ashley Leeds and Jeffrey M. Ritter and Sara McLaughlin Mitchell and Andrew G. Long},
#>   Journal = {International Interactions},
#>   Pages = {237--260},
#>   Title = {Alliance Treaty Obligations and Provisions, 1815-1944},
#>   Volume = {28},
#>   Year = {2002},
#>   Keywords = {add_atop_alliance()}
#> }
#> 
#> 
#> @Book{gibler2009ima,
#>   Author = {Douglas M. Gibler},
#>   Publisher = {Washington DC: CQ Press},
#>   Title = {International Military Alliances, 1648-2008},
#>   Year = {2009},
#>   Keywords = {add_cow_alliance()}
#> }

This function might expand in complexity in future releases, but you can use it right now for finding appropriate citations. You an also scan the ps_bib data to see what is in there.

Issues/Requests

{peacesciencer} is already more than capable to meet a wide variety of needs in the peace science community. Users are free to raise an issue on the project’s Github if some feature is not performing as they think it should or if there are additions they would like to see.