Type: Package
Version: 0.2-2
Date: 2023-02-18
Title: Ecological Inference and Higher-Dimension Data Management
Author: Olivia Lau <olivia.lau@post.harvard.edu>, Ryan T. Moore <rtm@american.edu>, Michael Kellermann <mrkellermann@gmail.com>
Maintainer: Michael Kellermann <mrkellermann@gmail.com>
Depends: R (≥ 2.0.0)
Imports: MASS, coda, msm
Suggests: lattice
Description: Provides methods for analyzing R by C ecological contingency tables using the extreme case analysis, ecological regression, and Multinomial-Dirichlet ecological inference models. Also provides tools for manipulating higher-dimension data objects.
License: GPL-2 | GPL-3 | file LICENSE [expanded from: GPL (≥ 2) | file LICENSE]
URL: http://www.olivialau.org/software/
NeedsCompilation: yes
Packaged: 2023-02-18 22:43:58 UTC; Mike Kellermann
Repository: CRAN
Date/Publication: 2023-02-18 23:40:02 UTC

Deterministic bounds for units satisfying row thresholds

Description

Calculates the deterministic bounds on the proportion of row members within a specified column.

Usage

bounds(formula, data, rows, column, excluded = NULL, 
    threshold = 0.9, total = NULL)

Arguments

formula

a formula of the form cbind(col1, col2, ...) ~ cbind(row1, row2, ...). Column and row marginals must have the same total for each ecological unit.

data

a data frame containing the variables specified in formula and (optionally) total

rows

a character vector specifying the rows of interest

column

a character string specifying the column marginal of interest

excluded

an optional character string (or vector of character strings) specifying the columns to be excluded from the bounds calculation. For example, if the quantity of interest is Democratic share of the two-party vote, non-voters would be excluded.

threshold

the minimum proportion of the unit that row members must comprise for the bounds to be calculated for the unit. If threshold = 0, bounds will be calculated for all units.

total

if row and/or column marginals are given as proportions, total identifies the name of the variable in data containing the total number of individuals in each unit

Value

A list with elements

bounds

a list of deterministic bounds for all units in which row proportions meet the threshold

intersection

if the intersection of the deterministic bounding intervals is non-empty, the intersection is returned. Otherwise, NA is returned.

Author(s)

Ryan T. Moore <rtm@american.edu>

References

Otis Dudley Duncan and Beverley Davis. 1953. “An Alternative to Ecological Correlation.” American Sociological Review 18: 665-666.

See Also

plot.bounds

Unit-level coverage plots for beta parameters from MD EI model

Description

Generates a plot of central credible intervals for the unit-level beta parameters from the Multinomial-Dirichlet ecological inference model (see ei.MD.bayes).

Usage

cover.plot(object, row, column, x = NULL, CI = 0.95,
          medians = TRUE, col = NULL, ylim = c(0,1), 
          ylab, lty = par("lty"), lwd = par("lwd"), ...)

Arguments

object

output from ei.MD.bayes

row

a character string specifying the row marginal of interest

column

a character string specifying the column marginal of interest

x

an optional covariate to index the units along the x-axis

CI

a fraction between 0 and 1 (defaults to 0.95), specifying the coverage of the central credible interval to be plotted for each unit

medians

a logical value specifying whether to plot the median (defaults to TRUE). If medians = FALSE, the medians are not plotted.

col

an optional vector of colors to be passed to plot and segments. If col is of length two, then the first color is used for plot and the second for segments.

ylim

an optional range for the y-axis (defaults to c(0,1)).

ylab

an optional label for the y-axis (defaults to Proportion of row in column).

lty

an optional line type passed to segments.

lwd

an optional line width argument passed to segments.

...

additional arguments passed to plot.

Value

A plot with vertical intervals indicating the central credible intervals for each ecological unit.

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu>

See Also

plot, segments, par

Density plots for population level parameters

Description

Generates a density plot for population level quantities of interest output by lambda.MD, lambda.reg, and lambda.reg.bayes. For the Bayesian methods, densityplot plots the kernel density for the draws. For the frequentist lambda.reg method, densityplot plots the canonical Normal density conditional on the mean and standard error output by lambda.reg.

Usage

## S3 method for class 'lambdaMD'
densityplot(x, by = "column", col, xlim, ylim,
             main = "", sub = NULL, xlab, ylab,
             lty = par("lty"), lwd = par("lwd"), ...)
## S3 method for class 'lambdaRegBayes'
densityplot(x, by = "column", col, xlim, ylim,
             main = "", sub = NULL, xlab, ylab,
             lty = par("lty"), lwd = par("lwd"), ...)
## S3 method for class 'lambdaReg'
densityplot(x, by = "column", col, xlim, ylim,
             main = "", sub = NULL, xlab, ylab,
             lty = par("lty"), lwd = par("lwd"), ...)

Arguments

x

output from lambda.MD, lambda.reg, or lambda.reg.bayes.

by

character string (defaulting to "column") specifying whether to panel the density plot by "row" or "column" marginal.

col

an optional vector of colors, with length corresponding to the number of marginals selected in by. Defaults to rainbow.

xlim, ylim

optional limits for the x-axis and y-axis, passed to plot.

main, sub

optional title and subtitle, passed to plot.

xlab, ylab

optional labels for the x- and y-axes, passed to plot.

lty, lwd

optional arguments for line type and line width, passed to lines and plot. If either lty or lwd are vectors, it must correspond to the number of row or column marginals selected.

...

additional arguments passed to par.

Value

A plot with density lines for the selected margin (row or column).

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu>

See Also

plot, segments, par

Multinomial Dirichlet model for Ecological Inference in RxC tables

Description

Implements a version of the hierarchical model suggested in Rosen et al. (2001)

Usage

ei.MD.bayes(formula, covariate = NULL, total = NULL, data, 
            lambda1 = 4, lambda2 = 2, covariate.prior.list = NULL,
            tune.list = NULL, start.list = NULL, sample = 1000, thin = 1, 
            burnin = 1000, verbose = 0, ret.beta = 'r', 
            ret.mcmc = TRUE, usrfun = NULL)

Arguments

formula

A formula of the form cbind(col1, col2, ...) ~ cbind(row1, row2, ...). Column and row marginals must have the same totals.

covariate

An optional formula of the form ~ covariate. The default is covariate = NULL, which fits the model without a covariate.

total

if row and/or column marginals are given as proportions, total identifies the name of the variable in data containing the total number of individuals in each unit

data

A data frame containing the variables specified in formula and total

lambda1

The shape parameter for the gamma prior (defaults to 4)

lambda2

The rate parameter for the gamma prior (defaults to 2)

covariate.prior.list

a list containing the parameters for normal prior distributions on delta and gamma for model with covariate. See ‘details’ for more information.

tune.list

A list containing tuning parameters for each block of parameters. See ‘details’ for more information. Typically, this will be a list generated by tuneMD. The default is NULL, in which case fixed tuning parameters are used.

start.list

A list containing starting values for each block of parameters. See ‘details’ for more information. The default is start.list = NULL, which generates appropriate random starting values.

sample

Number of draws to be saved from chain and returned as output from the function (defaults to 1000). The total length of the chain is sample*thin + burnin.

thin

an integer specifying the thinning interval for posterior draws (defaults to 1, but most problems will require a much larger thinning interval).

burnin

integer specifying the number of initial iterations to be discarded (defaults to 1000, but most problems will require a longer burnin).

verbose

an integer specifying whether the progress of the sampler is printed to the screen (defaults to 0). If verbose is greater than 0, the iteration number is printed to the screen every verboseth iteration.

ret.beta

A character indicating how the posterior draws of beta should be handled: 'r'eturn as an R object, 's'ave as .txt.gz files, 'd'iscard (defaults to r).

ret.mcmc

A logical value indicating how the samples from the posterior should be returned. If TRUE (default), samples are returned as coda mcmc objects. If FALSE, samples are returned as arrays.

usrfun

the name of an optional a user-defined function to obtain quantities of interest while drawing from the MCMC chain (defaults to NULL).

Details

ei.MD.bayes implements a version of the hierarchical Multinomial-Dirichlet model for ecological inference in R \times C tables suggested by Rosen et al. (2001).

Let r = 1, \ldots, R index rows, C = 1, \ldots, C index columns, and i = 1, \ldots, n index units. Let N_{\cdot ci} be the marginal count for column c in unit i and X_{ri} be the marginal proportion for row r in unit i. Finally, let \beta_{rci} be the proportion of row r in column c for unit i.

The first stage of the model assumes that the vector of column marginal counts in unit i follows a Multinomial distribution of the form:

(N_{\cdot 1i}, \ldots, N_{\cdot Ci}) {\sim} {\rm Multinomial}(N_i,\sum_{r=1}^R \beta_{r1i}X_{ri}, \dots, \sum_{r=1}^R \beta_{rCi}X_{ri})

The second stage of the model assumes that the vector of \beta for row r in unit i follows a Dirichlet distribution with C parameters. The model may be fit with or without a covariate.

If the model is fit without a covariate, the distribution of the vector \beta_{ri} is :

(\beta_{r1i}, \dots, \beta_{rCi}) {\sim} {\rm Dirichlet}(\alpha_{r1}, \dots, \alpha_{rC})

In this case, the prior on each \alpha_{rc} is assumed to be:

\alpha_{rc} \sim {\rm Gamma}(\lambda_1, \lambda_2)

If the model is fit with a covariate, the distribution of the vector \beta_{ri} is :

(\beta_{r1i}, \dots, \beta_{rCi}) {\sim} {\rm Dirichlet}(d_r\exp(\gamma_{r1} + \delta_{r1}Z_i), d_r\exp(\gamma_{r(C-1)} + \delta_{r(C-1)}Z_i), d_r)

The parameters \gamma_{rC} and \delta_{rC} are constrained to be zero for identification. (In this function, the last column entered in the formula is so constrained.)

Finally, the prior for d_r is:

d_r \sim {\rm Gamma}(\lambda_1, \lambda_2)

while \gamma_{rC} and \delta_{rC} are given improper uniform priors if covariate.prior.list = NULL or have independent normal priors of the form:

\delta_{rC} \sim {\rm N}(\mu_{\delta_{rC}}, \sigma_{\delta_{rC}}^2)

\gamma_{rC} \sim {\rm N}(\mu_{\gamma_{rC}}, \sigma_{\gamma_{rC}}^2)

If the user wishes to estimate the model with proper normal priors on \gamma_{rC} and \delta_{rC}, a list with four elements must be provided for covariate.prior.list:

Applying the model without a covariate is most reasonable in situations where one can think of individuals being randomly assigned to units, so that there are no aggregation or contextual effects. When this assumption is not reasonable, including an appropriate covariate may improve inferences; note, however, that there is typically little information in the data about the relationship of any given covariate to the unit parameters, which can lead to extremely slow mixing of the MCMC chains and difficulty in assessing convergence.

Because the conditional distributions are non-standard, draws from the posterior are obtained by using a Metropolis-within-Gibbs algorithm. The proposal density for each parameter is a univariate normal distribution centered at the current parameter value with standard deviation equal to the tuning constant; the only exception is for draws of \gamma_{rc} and \delta_{rc}, which use a bivariate normal proposal with covariance zero.

The function will accept user-specified starting values as an argument. If the model includes a covariate, the starting values must be a list with the following elements, in this order:

If there is no covariate, the starting values must be a list with the following elements:

The function will accept user-specified tuning parameters as an argument. The tuning parameters define the standard deviation of the normal distribution used to generate candidate values for each parameter. For the model with a covariate, a bivariate normal distribution is used to generate proposals; the covariance of these normal distributions is fixed at zero. If the model includes a covariate, the tuning parameters must be a list with the following elements, in this order:

If there is no covariate, the tuning parameters are a list with the following elements:

Value

A list containing

draws

A list containing samples from the posterior distribution of the parameters. If a covariate is included in the model, the list contains:

  • Dr Posterior draws for Dr parameters as an R \timessample matrix. If ret.mcmc = TRUE, Dr is an mcmc object.

  • Beta Posterior draws for beta parameters. Only returned if ret.beta = TRUE. If ret.mcmc = TRUE, a (R * C * units) \times sample matrix saved as an mcmc object. Otherwise, a R \times C \times units \times sample array

  • Gamma Posterior draws for gamma parameters. If ret.mcmc = TRUE, a (R * (C - 1)) \times sample matrix saved as an mcmc object. Otherwise, a R \times (C - 1) \times sample array

  • Delta Posterior draws for delta parameters. If ret.mcmc = TRUE, a (R * (C - 1)) \times sample matrix saved as an mcmc object. Otherwise, a R \times(C - 1) \times sample array

  • Cell.count Posterior draws for the cell counts, summed across units. If ret.mcmc = TRUE, a (R * C) \times sample matrix saved as an mcmc object. Otherwise, a R \times C \times sample array

If the model is fit without a covariate, the list includes:

  • Alpha Posterior draws for alpha parameters. If ret.mcmc = TRUE, a (R * C) \times sample matrix saved as an mcmc object. Otherwise, a R \times C \times sample array

  • Beta Posterior draws for beta parameters. If ret.mcmc = TRUE, a (R * C * units) \times sample matrix saved as an mcmc object. Otherwise, a R \times C \times units \times sample array

  • Cell.count Posterior draws for the cell counts, summed across units. If ret.mcmc = TRUE, a (R * C) \times sample matrix saved as anmcmc object. Otherwise, a R \times C \times sample array

acc.ratios

A list containing acceptance ratios for the parameters. If the model includes a covariate, the list includes:

  • dr.acc A vector of acceptance ratios for Dr draws

  • beta.acc A vector of acceptance ratios for Beta draws

  • gamma.acc A vector of acceptance ratios for Gamma and Delta draws

If the model is fit without a covariate , the list includes:

  • alpha.acc A vector of acceptance ratios for Alpha draws

  • beta.acc A vector of acceptance ratios for Beta draws

usrfun

Output from the optional usrfn

call

Call to ei.MD.bayes

Author(s)

Michael Kellermann <mrkellermann@gmail.com> and Olivia Lau <olivia.lau@post.harvard.edu>

References

Martyn Plummer, Nicky Best, Kate Cowles, and Karen Vines. 2002. Output Analysis and Diagnostics for MCMC (CODA). https://CRAN.R-project.org/package=coda.

Ori Rosen, Wenxin Jiang, Gary King, and Martin A. Tanner. 2001. “Bayesian and Frequentist Inference for Ecological Inference: The R \times (C-1) Case.” Statistica Neerlandica 55: 134-156.

See Also

lambda.MD, cover.plot, density.plot, tuneMD, mergeMD

Ecological regression

Description

Estimate an ecological regression using least squares.

Usage

ei.reg(formula, data, ...)

Arguments

formula

An R formula object of the form cbind(c1, c2, ...) ~ cbind(r1, r2, ...)

data

data frame containing the variables specified in formula

...

Additional arguments passed to lm.

Details

For i \in 1,\ldots,C, C regressions of the form c_i ~ cbind(r1, r2, ...) are performed.

These regressions make use of the accounting identities and the constancy assumption, that \beta_{rci} = \beta_{rc} for all i.

The accounting identities include

Then regressing

T_{ci} = \beta_{rc} X_{ri} + \epsilon_{ci}

for c = 1,\dots,C recovers the population parameters \beta_{rc} when the standard linear regression assumptions apply, including E[\epsilon_{ci}] = 0 and Var[\epsilon_{ci}] = \sigma_c^2 for all i.

Value

A list containing

call

the call to ei.reg

coefficients

an R \times C matrix of estimated population cell fractions

se

an R \times C matrix of standard errors for coefficients.

cov.matrices

A list of the C scaled variance-covariance matrices for each of the ecological regressions

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu> and Ryan T. Moore <rtm@american.edu>

References

Leo Goodman. 1953. “Ecological Regressions and the Behavior of Individuals.” American Sociological Review 18:663–664.

Ecological regression using Bayesian Normal regression

Description

Estimate an ecological regression using Bayesian normal regression.

Usage

ei.reg.bayes(formula, data, sample = 1000, weights = NULL, truncate=FALSE)

Arguments

formula

An R formula object of the form cbind(c1, c2, ...) ~ cbind(r1, r2, ...)

data

data frame containing the variables specified in formula

sample

number of draws from the posterior

weights

a vector of weights

truncate

if TRUE, imposes a proper uniform prior on the unit hypercube for the coefficients; if FALSE, an improper uniform prior is assumed

Details

For i \in 1,\ldots,C, C Bayesian regressions of the form c_i ~ cbind(r1, r2, ...) are performed. See the documentation for ei.reg for the accounting identities and constancy assumption underlying this Bayesian linear model.

The sampling density is given by

y|\beta, \sigma^2, X \sim N(X\beta, \sigma^2 I)

The improper prior is p(\beta,\sigma^2|X)\propto \sigma^{-2}.

The proper prior is p(\beta, \sigma^2|x) \propto I(\beta \in [0,1])\times \sigma^{-2}.

Value

A list containing

call

the call to ei.reg.bayes

draws

A, R \times C \times sample array containing posterior draws for each population cell fraction

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu> and Ryan T. Moore <rtm@american.edu>

References

Leo Goodman. 1953. “Ecological Regressions and the Behavior of Individuals.” American Sociological Review 18:663–664.

Calculate shares using data from MD model

Description

Calculates the population share of row members in a particular column as a proportion of the total number of row members in the selected subset of columns.

Usage

lambda.MD(object, columns, ret.mcmc = TRUE)

Arguments

object

an R object of class eiMD, output from ei.MD.bayes

columns

a character vector of column names to be included in calculating the shares

ret.mcmc

a logical value indicating how the samples from the posterior should be returned. If TRUE (default), samples are returned as mcmc objects. If FALSE, samples are returned as arrays.

Details

This function allows users to define subpopulations within the data and calculate the proportion of individuals within each of the columns that defines that subpopulation. For example, if the model includes the groups Democrat, Republican, and Unaffiliated, the argument columns = c(``Democrat", ``Republican") will calculate the two-party shares of Democrats and Republicans for each row.

Value

Returns either a ((R * included columns) \times samples) matrix as an mcmc object or a (R \times included columns \times samples) array.

Author(s)

Michael Kellermann <mrkellermann@gmail.com> and Olivia Lau <olivia.lau@post.harvard.edu>

See Also

ei.MD.bayes

Calculate shares using data from regression model

Description

Calculates the population share of row members in a particular column

Usage

lambda.reg(object, columns)

Arguments

object

An R object of class eiReg, the output from ei.reg

columns

a character vector of column names to be included in calculating the shares

Details

Standard errors are calculated using the delta method as implemented in the library msm. The arguments passed to deltamethod in msm include

cov = Var(\hat{\beta}_{r1}) 0 0 \ldots
0 Var(\hat{\beta}_{r2}) 0 \ldots
0 0 Var(\hat{\beta}_{r3}) \ldots
\vdots \vdots \vdots \ddots

Value

Returns a list with the following elements

call

the call to lambda.reg

lambda

an R \times k matrix where k is the number of columns included in the share calculation

se

standard errors calculated using the delta method as implemented in the library msm

Author(s)

Ryan T. Moore <rtm@american.edu>

See Also

ei.reg

Calculate shares using data from Bayesian regression model

Description

Calculates the population share of row members in selected columns

Usage

lambda.reg.bayes(object, columns, ret.mcmc = TRUE)

Arguments

object

An R object of class eiRegBayes, the output from ei.reg.bayes

columns

a character vector indicating which column marginals to be included in calculating the shares

ret.mcmc

If TRUE, posterior shares are returned as an mcmc object.

Value

If ret.mcmc = TRUE, draws are returned as an mcmc object with dimensions sample \times C. If ret.mcmc = FALSE, draws are returned as an array with dimensions R \times C \times samples array.

Author(s)

Ryan T. Moore <rtm@american.edu>

See Also

ei.reg.bayes

Combine output from multiple eiMD objects

Description

Allows users to combine output from several chains output by ei.MD.bayes

Usage

mergeMD(list, discard = 0)

Arguments

list

A list containing the names of multiple eiMD objects generated from the same model.

discard

The number of draws to discard from the beginning of each chain. Default is to retain all draws.

Value

Returns an eiMD object of the same format as the input.

Author(s)

Michael Kellermann <mrkellermann@gmail.com>

References

Martyn Plummer, Nicky Best, Kate Cowles, and Karen Vines. 2002. Output Analysis and Diagnostics for MCMC (CODA). https://CRAN.R-project.org/package=coda.

Ori Rosen, Wenxin Jiang, Gary King, and Martin A. Tanner. 2001. “Bayesian and Frequentist Inference for Ecological Inference: The R \times C Case.” Statistica Neerlandica 55: 134-156.

See Also

ei.MD.bayes

Plot of deterministic bounds for units satisfying row thresholds

Description

Plots the deterministic bounds on the proportion of row members within a specified column.

Usage

## S3 method for class 'bounds'
plot(x, row, column, labels = TRUE, order = NULL,
    intersection = TRUE, xlab, ylab, col = par("fg"), 
    lty = par("lty"), lwd = par("lwd"), ...)

Arguments

x

output from bounds

row

a character string specifying the row of interest

column

a character string specifying the column of interest

labels

a logical toggle specifying whether precinct labels should be printed above interval bounds

order

an optional vector of values between 0 and 1 specifying the order (left-to-right) in which interval bounds are plotted

intersection

a logical toggle specifying whether the intersection of all plotted bounds (if it exists) should be plotted

xlab, ylab, ...

additional arguments passed to plot

col, lty, lwd

additional arguments passed to segments

Value

A plot with vertical intervals indicating the deterministic bounds on the quantity of interest, and (optionally) a single horizontal interval indicating the intersection of these unit bounds.

Author(s)

Ryan T. Moore <rtm@american.edu>

See Also

bounds

Function to read in eiMD parameter chains saved to disk

Description

In ei.MD.bayes, users have the option to save parameter chains for the unit-level betas to disk rather than returning them to the workspace. This function reconstructs the parameter chains by reading them back into R and producing either an array or an mcmc object.

Usage

read.betas(rows, columns, units, dir = NULL, ret.mcmc = TRUE)

Arguments

rows

a character vector of the row marginals to be read back in

columns

a character vector of the column marginals to be read back in

units

a character of numeric vector with the units to be read back in

dir

an optional character string identifying the directory in which parameter chains are stored (defaults to getwd)

ret.mcmc

a logical value specifying whether to return the parameters as an mcmc object (defaults to TRUE)

Value

If ret.mcmc = TRUE, an mcmc object with row names corresponding to the parameter chains. If ret.mcmc = FALSE, an array with dimensions named according to the selected rows, columns, and units.

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu>

See Also

ei.MD.bayes,mcmc

Redistricting Monte-Carlo data

Description

Precinct-level observations for a hypothetical jurisdiction with four proposed districts.

Usage

data(redistrict)

Format

A table containing 150 observations and 9 variables:

precinct

precinct identifier

district

proposed district number

avg.age

average age

per.own

percent homeowners

black

number of black voting age persons

white

number of white voting age persons

hispanic

number of hispanic voting age persons

total

total number of voting age persons

dem

Number of votes for the Democratic candidate

rep

Number of votes for the Republican candidate

no.vote

Number of non voters

Source

Daniel James Greiner

Party registration in south-east North Carolina

Description

Registration data for White, Black, and Native American voters in eight counties of south-eastern North Carolina in 2001.

Usage

data(senc)

Format

A table containing 212 observations and 18 variables:

county

county name

precinct

precinct name

total

number of registered voters in precinct

white

number of White registered voters

black

number of Black registered voters

natam

number of Native American registered voters

dem

number of registered Democrats

rep

number of registered Republicans

other

number of registered voters without major party affiliation

whdem

number of White registered Democrats

whrep

number of White registered Republicans

whoth

number of White registered voters without major party affiliation

bldem

number of Black registered Democrats

blrep

number of Black registered Republicans

bloth

number of Black registered voters without major party affiliation

natamdem

number of Native American registered Democrats

natamrep

number of Native American registered Republicans

natamoth

number of Native American registered voters without major party affiliation

Source

Excerpted from North Carolina General Assembly 2001 redistricting data, https://www.ncleg.gov/Redistricting/BaseData2001

Tuning parameters for alpha hyperpriors in RxC EI model

Description

Tuning parameters for hyperpriors in RxC EI model

Usage

data(tuneA)

Format

A table containing 3 rows and 3 columns.

Tuning parameters for the precinct level parameters in the RxC EI model

Description

A vector containing tuning parameters for the precinct level parameters in the RxC EI model.

Usage

data(tuneB)

Format

A vector of length 3 x 2 x 150 containing the precinct level tuning parameters for the redistricting sample data.

Examples

data(tuneB)
tuneB <- array(tuneB[[1]], dim = c(3, 2, 150))

Generate tuning parameters for MD model

Description

An adaptive algorithm to generate tuning parameters for the MCMC algorithm implemented in ei.MD.bayes. Since we are drawing each parameter one at a time, target acceptance rates are between 0.4 to 0.6.

Usage

tuneMD(formula, covariate = NULL, data, ntunes = 10, 
    totaldraws = 10000, ...)

Arguments

formula

A formula of the form cbind(col1, col2, ...) ~ cbind(row1, row2, ...) with rows as the predictor and columns as the response

covariate

An R formula for the optional covariate in the form ~ x

data

data frame containing the variables specified in formula and covariate

ntunes

number of times to iterate the tuning algorithm

totaldraws

number of iterations for each tuning run

...

additional arguments passed to ei.MD.bayes

Value

A list containing matrices of tuning parameters.

Author(s)

Olivia Lau <olivia.lau@post.harvard.edu>

See Also

ei.MD.bayes

mirror server hosted at Truenetwork, Russian Federation.