This version has a few odds and ends and a version accelerated to 1.0.0 to coincide with the acceptance of a manuscript describing this package in Conflict Management and Peace Science.
add_cow_mids() and add_gml_mids() about the default case-exclusion rules for dyadic analyses.add_fpsim() and download_extdata() for more.cow_nmc) to version 6.0.cow_ddy and gw_ddy, since it’s 2022 now.grh_arms_races for arms race data.ps_bib.ps_cite() to no longer depend on {bib2df}. With it, remove {bib2df} as a package dependency.This is a leader-intensive update. Additions/new features include:
LEAD: a data set on select leader experience and attribute descriptions (LEAD).lwuf: a data set from Carter and Smith (2020) measuring leader willingness to use force.download_extdata() now also downloads directed leader dyad-year data for all directed leader dyad-years from 1870 to 2015. The data come in two forms. One is all possible (inter-state, leader-overlap) dyadic leader pairings from 1870 to 2015, as presented in the Archigos data. The second standardizes these data to the CoW state system data. It also downloads directed dyad-yer data of chance-corrected measures of dyadic foreign policy similarity. Type ?download_extdata() in the R session after updating for more information.create_leaderdyadyears(): create leader dyad-year data (via archigos) for all leader dyad-year combinations from 1870 to 2015. This function leans on remote data available for download in download_extdata(). The function returns either all possible Gleditsch-Ward dyadic leader pairings (standardized to Gleditsch-Ward state system dates) or standardizes the Archigos data to Correlates of War state system member data.gml_part: a data set on participant summary information from the GML MID data.add_fpsim() adds estimates of dyadic foreign policy similarity to your dyad-year and leader-dyad-year data.add_gml_mids() now works with leader-year data. See documentation for more.add_spells() will do more general peace spell calculations, though this package will now depend on v. 1.3.0 of {stevemisc}.add_lwuf() adds estimates of leader willingness to use force to leader-year and leader-dyad-year analyses.add_lead() adds some leader attributes to leader-year or leader-dyad-year analyses.gml_mid_dirleaderdisps and gml_mid_ddlydisps include data on leader dyads in conflict.declare_attributes() allows users to declare package-specific attributes to outside data they bring to the R session. This allows the user to use many of the shortcuts in this package to their data.ps_version() will display version information about the data included in this package.false_cow_dyads and false_gw_dyads list a handful of dyads that appear in the same year, but never on the same day in the year. They’re used for anti-joining in create_dyadyears().leader_codes matches, as well as one can, leader/observation codes in Archigos 4.1, Archigos 2.9, and the LEAD data.Other changes include:
add_creg_fractionalization(), add_rugged_terrain(), add_democracy(), and more. Some functions, like add_gml_mids(), can be anchored explicitly to leader-level units of analyses. Several other functions (e.g. add_sdp_gdp() and others like it) are fundamentally still state-year or dyad-year functions the extent to which leaders are still nested in states and leader-dyads are just a derivation of inter-state dyads (i.e. leaders are nested in states). Use with some caution with that in mind, though these functions are still useful for leader-level analyses that are interested in some of these covariates for making state-to-state comparisons on top of leader-level analyses.cow_sdp_gdp and gw_sdp_gdp are now rounded to three decimal points. These “economic” data are routinely the biggest in the package, and it’s because of the decimal points. The justification for this is these data are estimated/simulated anyways and the information loss is at the 1/1000th decimal point. This procedure basically cuts the size of the data to be less than 25% of its original size. This is a huge reduction in disk space.leader_year) is now an attribute. Where appropriate, leader-year and state-year functions are treated the same. For example, the GDP data contained in add_sdp_gdp() are fundamentally state-year, but perhaps a leader-year control of interest even as leaders are nested in states.leader_dyad_year) is now an attribute. Where appropriate, leader-year and leader-dyad-year functions are treated the same. Where possible/tractable, leader-dyad-year data are explicitly anchored to leader-level units of analysis (e.g. add_gml_mids(), prominently).create_leaderyears() works on the observation ID and not the leader ID. Observation IDs are unique to each leader-period, but leader IDs are not.add_peace_years() has been superseded by add_spells(). The “superseded” language communicates the function will still work as it has for previous releases, for data types supported by the function, though its development will stop. add_spells() promises to have greater functionality. Communicating these changes via badges brings in {lifecycle} as an imported package.archigos data’s ccode has been renamed to gwcode to reflect these are actually Gleditsch-Ward state codes.create_leaderyears(), create_dyadyears(), and create_stateyears() have a built-in argument for subsetting the years returned by the function (subset_years). Use to your discretion.cow_mindist and gw_mindist data, given new {cshapes} updates. Data now run from 1886 to 2019.cow_capitals and gw_capitals data. Remove capitals data for redundancy. Both capital data sets extended to 2020.add_capital_distance() now works with Correlates of War and Gleditsch-Ward data, both dyad-year and state-year.add_atop_alliance() that added 0s to years after the ATOP domain. Thanks to @joshuaalley for pointing this out.{lubridate}. Earlier versions of this package uncritically leaned on {lubridate} for functions I could either write myself (i.e. year(), which is now .pshf_year()) or were already duplicated in base functions in R (i.e. ymd(), a wrapper for as.Date() as I use it). With it, {lubridate} is no longer a requirement for this package.gml_mid_disps.ps_conflict_type attributes for some conflict data (e.g. CoW-MID, GML).add_cow_mids() and add_gml_mids().whittle_conflicts_onsets(), whittle_conflicts_fatality(), whittle_conflicts_hostility(), whittle_conflicts_duration(), whittle_conflicts_reciprocation(), whittle_conflicts_startmonth(), and whittle_conflicts_jds(). These are admittedly gnarly function names to type out if you don’t have an IDE like RStudio to assist you. Thus, these respective functions come with shortcuts of wc_ (e.g. wc_onsets()). Check the package documentation to see what these are.show_duplicates() as a diagnostic tool. This function is useful for users who want to do some advanced stuff with data created in {peacesciencer} (e.g. merging in custom data) and want to see if they botched a merge by creating duplicate observations.archigos data frame now includes some more information about leaders (e.g. name, gender, and year of birth). create_leaderyears() will create leader-year data as well. The next update will expand more on leader functions. For now, the ensuing output from this function is treated as synonymous to state-year data.download_extdata() as a way of side-stepping package space limitations. Some files will have to be stored remotely and then loaded at the user’s discretion, which is what this function will do. This was largely brought on by the CoW trade data (cow_trade_ndy), which is also removed in this update. More data may have to be moved remotely in the future, though this package will endeavor to keep this to a bare minimum. Importantly, download_extdata() keeps inventory of what it is downloading. Data information is included therein. This develop does implicitly assume that the directory in which this package is installed is writable by the user. For like 99% of users, this shouldn’t be a problem (and executing .libPaths() should confirm that). Do reach out if it is.creg data frame.{bib2df} as package dependency. Force upgraded version of {stevemisc} as package dependency.LazyDataCompression: xz to DESCRIPTIONcreate_dyadyears(), create_statedays(), and create_stateyears().add_cow_wars() function.add_ccode_to_gw(). Fix underlying data for cow_gw_years. Add gw_cow_years.add_strategic_rivalries().add_ucdp_acd(). With it, update ucdp_acd data for functionality.add_rugged_terrain().cow_trade_ndy data. With it, update add_cow_trade() function for speed upgrades.cow_war_inter, cow_war_intra data.creg data.gwcode_democracy data. add_democracy() now takes CoW or G-W data.hief data.rugged data.td_rivalries data.ps_cite() and ps_bib. Remove citations.cow_sdp_gdp and gw_sdp_gdp. Update add_sdp_gdp() to reflect this change.add_contiguity() for why non-contiguous dyads are 0s and not 6s. Briefly: I don’t think of these data as ordinal and I don’t encourage the researcher to think of them as ordinal either.add_peace_years(), which now works with UCDP conflict data.filter_prd() is now a shortcut for add_contiguity() and/or add_cow_majors() if it is called near the top of the pipe (i.e. before the user executed the commands required to calculate politically relevant dyad status).add_nmc() (and cow_nmc). -9 is now explicitly NA, as it should have been.add_peace_years(), which can now be used anywhere in the chain of commands (instead of just last). add_peace_years() is also a bit more “general” and includes a derivation of the sbtscs() function from the {stevemisc} package.add_minimum_distance() and add_sdp_gdp(). These merges now lean on system attributes declared in create_dyadyears() and create_stateyears().add_mids(). Use add_gml_mids() instead.cow_ddy and its documentation.cow_mid_ddydisps, cow_mid_dirdisps, and cow_mid_disps data.gml_mid_ddydisps data.add_cow_mids() function.add_gml_mids() function.add_peace_years() function.add_mids(). Use add_gml_mids() instead.ps_data_type == "dyad_year" as attribute to cow_mid_dirdisps and gml_dirdisp. This allows {peacesciencer} functions for these data sets.atop_alliance (directed dyad-year) dataarchigos datacow_sdp_gdp and gw_sdp_gdp dataucdp_acd dataucdp_onsets dataadd_archigos() functionadd_atop_alliance() functionadd_minimum_distance() function. For now, this has no default. You must specify “gw” or “cow”.add_sdp_gdp() function. For now, this has no default. You must specify “gw” or “cow”.add_ucdp_onset() function.filter_prd(). Past versions of my code had treated non-contiguous dyads as 6s. Here, they’re 0s. That created situations where I wasn’t filtering anything at all with this function.cow_alliance to distinguish from ATOP equivalents.create_statedays() function.