The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
county_turnout_panel (random forest via
ranger), adds a summary table of method assumptions and
coverage guarantees, and a new appendix evaluating the example with
interval_score() and interval_width().
ranger added to Suggests.importFrom("stats", "lm.fit") to the
NAMESPACEcounty_turnout_panel gains 16 columns (now 38 total):
total_votes and cvap — the turnout
numerator and denominator, so turnout and every share is
reproducible from the shipped data.votes_dem, votes_rep,
votes_other and the matching dem_share,
rep_share, other_share — presidential votes
and vote shares by party (Libertarian/Green pooled into “other” for a
consistent panel).election_year: turnout_county_mean /
turnout_electorate and
{dem,rep,other}_share_county_mean /
_electorate — the unweighted county-wise mean and the
pooled electorate-wide (“true”) turnout and vote shares for each
election, computed over the full electorate.county_turnout_panel now excludes the 25
micro-county-years whose recorded presidential ballots exceeded the
noisy 5-year ACS CVAP estimate (turnout > 1), leaving 15,543 rows.
Rows with an unknown (NA) turnout are still retained. The
filtering is performed by the build pipeline in
data-raw/.county_turnout_panel: a tidy five-wave
county-level panel of U.S. presidential-election turnout and ACS
demographic context for the 2008, 2012, 2016, 2020 and 2024 elections
(15,568 rows). Turnout is votes cast divided by Citizen Voting-Age
Population (CVAP). All demographic covariates are re-derived directly
from the U.S. Census Bureau API (not IPUMS NHGIS, to comply with NHGIS’s
non-redistribution terms) and match the vintage of each election. Ships
with a full codebook at
system.file("extdata", "county_turnout_panel_codebook.md", package = "pintervals")
and a reproducible build pipeline in data-raw/.county_turnout (the original single-election 2016
cross-section) is superseded by
county_turnout_panel but retained for
backward compatibility; it is still used by the package vignette and
examples. Its data are unchanged.?county_turnout previously stated
that turnout was “total votes cast divided by total
population”. It is in fact votes divided by the Citizen Voting-Age
Population (CVAP); the help page has been corrected.
county_turnout_panel$turnout uses this same definition and
reproduces the 2016 county_turnout values for 3,103 of the
3,107 shared counties. The Details section of
?county_turnout (the predicted_turnout
reproduction code) also now renders — it was previously dropped because
of an unescaped % in the Rd source.These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.