All data are stored in an object of class judgeit . If desired,
the user can access each component within. Here is a list of
components and their attributes:
- covars is a list of data frames comprising the
predictors for each election in the system. So
judgeit.object$covars[[25]] is a data frame with the
covariates from the 25th election.
- voteshare is a list of vectors comprising the vote
shares for each election in the system. So
judgeit.object$voteshare[[25]] is a vector of the results
of the 25th election.
- turnout,elgvotes,seats are lists of vectors
comprising the actual turnout, the number of eligible voters, and
the seats per district in the system for each election.
- fullrow is a list of vectors containing those rows
whose primary elements (covariates, vote shares, eligible and actual
voters and seats) contain complete data.
- uncL,uncLR,uncU,uncUR are the uncontested election
detection thresholds and imputations as listed above.
- svexpected.value.only is the value of expected.value.only
as given above.
- simulations is the number of simulations conducted by
JudgeIt during each analysis.
- weight is the option selected by the user to indicate
what weights should be used in the linear model, as described above.
- distweights is a list of the actual values of these
weights.
- covarsnew is a list of data frames of counterfactual or
future predictors as manipulated by the option
new.covariates. It must have the same data type in each column as covars though not necessarily the same number of rows.
- same.dists is a vector indicating whether the previous
election's district map is identical to the current one, as
described above.
- output contains the output of the last analytical
routine, and is displayed with the command
print(JudgeIt.object).
- outputyear, outputclass indicate the year and type of
the last analysis conducted. These are used mainly in
plot(judgeit.object).
- beta,vc are the estimates given by the linear model for
the system for the coefficients of the covariates and their
covariance matrix.
- sigma,lambda,sind,lind are, respectively, the mean and
year-by-year estimates of the standard error and systematic error
fraction of the system.
- years is a vector of the names of each election
variable in the inputted data frame list. In the case of
house6311, this is a vector of the election years between
1896 and 1992.
Gary King
2010-08-31