Skip to content

Releases: svmiller/peacesciencer

v1.1.0

24 Mar 14:16
Compare
Choose a tag to compare
  • Package now depends on version 1.6.0 of {stevemisc}. This comes as later versions of R balk at the use of order() on data frames.
  • Changes to documentation for clarity/presentation.
  • Changes to file compression, where appropriate.
  • download_extdata() will download some augmented dyadic minimum distance data for CoW and G-W states from 1886 to 2019.
  • Add tss_rivalries as updated rivalry data from 1494 to 2020.
  • Fixed a bug in atop_alliance and cow_alliance. It’s less of a peacesciencer bug, per se, but the raw data that were presented as directed were actually not directed at all. They are now.
  • Fixed a bug in add_cow_alliance() that did not honor the temporal bound of the alliance data (i.e. 1816-2012). My b.

v1.0.0

24 Mar 19:25
Compare
Choose a tag to compare

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.

  • Clarify add_cow_mids() and add_gml_mids() about the default case-exclusion rules for dyadic analyses.
  • Assorted documentation upgrades for clarity.
  • The dyadic foreign policy similarity data now have Tau-b estimates. These were generated by me, using valued alliance data, and should be used with some caution (or perhaps not used at all). Check the documentation for add_fpsim() and download_extdata() for more.
  • Update national material capabilities data (cow_nmc) to version 6.0.
  • Update cow_ddy and gw_ddy, since it's 2022 now.
  • Add grh_arms_races for arms race data.
  • Amend/update a few items in ps_bib.
  • Change ps_cite() to no longer depend on {bib2df}. With it, remove {bib2df} as a package dependency.

v0.7.0

11 Nov 17:03
Compare
Choose a tag to compare

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:

  • Several functions now have leader-year or leader-dyad-year corollaries. Examples here 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.
  • The data in 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" (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" (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.
  • The directed leader dyad-year data is indexed by observation ID and not leader ID, per se. This is because leaders can have multiple terms in office.
  • 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.
  • The "whittle" functions now work with leader-dyadic conflict data available in the package.
  • The package has started taking inventory of non-ASCII characters that appear in the package (e.g. spelling "Wuerttemberg" with the U-umlaut) and replacing them. This is not a functionality issue at all. The problem is CRAN raises a note for every single non-ASCII character it sees and I'd like to avoid the CRAN hammer if I can.
  • 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.