multinomRob: Multinomial Robust Estimation

In multinomRob: Robust Estimation of Overdispersed Multinomial Regression Models

Description

multinomRob fits the overdispersed multinomial regression model for grouped count data using the hyperbolic tangent (tanh) and least quartile difference (LQD) robust estimators.

Usage

  multinomRob(model, data, starting.values=NULL,  equality=NULL,
              genoud.parms=NULL, print.level=0, iter = FALSE,
              maxiter = 10, multinom.t=1, multinom.t.df=NA,
              MLEonly=FALSE)

Arguments

model	The regression model specification. This is a list of formulas, with one formula for each category of outcomes for which counts have been measured for each observation. For example, in the following, model=list(y1 ~ x1, y2 ~ x2, y3 ~ 0) the outcome variables containing counts are y1, y2 and y3, and the linear predictor for y1 is a coefficient times x1 plus a constant, the linear predictor for y2 is a coefficient times x2 plus a constant, and the linear predictor for y3 is zero. Each formula has the format countvar ~ RHS, where countvar is the name of a vector, in the dataframe referenced by the data argument, that gives the counts for all observations for one category. RHS denotes the righthand side of a formula using the usual syntax for formulas, where each variable in the formula is the name of a vector in the dataframe referenced by the data argument. For example, a RHS specification of var1 + var2*var3 would specify that the regressors are to be var1, var2, var3, the terms generated by the interaction var2:var3, and the constant. The set of outcome alternatives may be specified to vary over observations, by putting in a negative value for alternatives that do not exist for particular observations. If the value of an outcome variable is negative for an observation, then that outcome is considered not available for that observation. The predicted counts for that observation are defined only for the available observations and are based on the linear predictors for the available observations. The same set of coefficient parameter values are used for all observations. Any observation for which fewer than two outcomes are available is omitted. Observations with missing data (NA) in any outcome variable or regressor are omitted (listwise deletion). In a model that has the same regressors for every category, except for one category for which there are no regressors in order to identify the model (the reference category), the RHS specification must be given for all the categories except the reference category. The formula for the reference category must include a RHS specification that explicitly omits the constant, e.g., countvar ~ -1 or countvar ~ 0. The number of coefficient parameters to be estimated equals the number of terms generated by all the formulas, subject to equality constraints that may be specified using the equality argument.
data	The dataframe that contains all the variables referenced in the model argument, which are the data to be analyzed.
starting.values	Starting values for the regression coefficient parameters, as a vector. The parameter ordering matches the ordering of the formulas in the model argument: parameters for the terms in the first formula appear first, then come parameters for the terms in the second formula, etc. In practice it will usually be better to start by letting multinomRob find starting values by using the multinom.t option, then using the results from one run as starting values for a subsequent run done with, perhaps, a larger population of operators for rgenoud.
equality	List of equality constraints. This is a list of lists of formulas. Each formula has the same format as in the model specification, and must include only a subset of the outcomes and regressors used in the model specification formulas. All the coefficients specified by the formulas in each list will be constrained to have the same value during estimation. For example, in the following, multinomRob(model=list(y1 ~ x1, y2 ~ x2, y3 ~ 0), data=dtf, equality=list(list(y1 ~ x1 + 0, y2 ~ x2 + 0)) ); the model to be estimated is list(y1 ~ x1, y2 ~ x2, y3 ~ 0) and the coefficients of x1 and x2 are constrained equal by equality=list(list(y1 ~ x1 + 0, y2 ~ x2 + 0)) In the equality formulas it is necessary to say + 0 so the intercepts are not involved in the constraints. If a parameter occurs in two different lists in the equality= argument, then all the parameters in the two lists are constrained to be equal to one another. In the output this is described as consolidating the lists.
genoud.parms	List of named arguments used to control the rgenoud optimizer, which is used to compute the LQD estimator.
print.level	Specify 0 for minimal printing, 1 to print more detailed information about LQD and other intermediate computations, 2 to print details about the tanh computations, or 3 to print details about starting values computations.
iter	TRUE means to iterate between LQD and tanh estimation steps until either the algorithm converges, the number of iterations specified by the maxiter argument is reached, or if an LQD step occurs that produces a larger value than the previous step did for the overdispersion scale parameter. This option is often improves the fit of the model.
maxiter	The maximum number of iterations to be done between LQD and tanh estimation steps.
multinom.t	1 means use the multinomial multivariate-t model to compute starting values for the coefficient parameters. But if the MNL results are better (as judged by the LQD fit), MNL values will be used instead. 0 means use nonrobust maximum likelihood estimates for a multinomial regression model. 2 forces the use of the multivariate-t model for starting values even if the MNL estimates provide better starting values for the LQD. Note that with multinom.t=1 or multinom.t=2, multivariate-t starting values will not be used if the model cannot generate valid standard errors. To force the use of multivariate-t estimates even in this circumstance, see the multinom.t.df argument. If the starting.values argument is not NULL, the starting values given in that argument are used and the multinom.t argument is ignored. Multinomial multivariate-t starting values are not available when the number of outcome alternatives varies over the observations.
multinom.t.df	NA means that the degrees of freedom (DF) for the multivariate-t model (when used) should be estimated. If multinom.t.df is a number, that number will be used for the degrees of freedom and the DF will not be estimated. Only a positive number should be used. Setting multinom.t.df to a number also implies that, if multinom.t=1 or multinom.t=2, the multivariate-t starting values will be used (depending on the comparison with the MNL estimates if multinom.t=1 is set) even if the standard errors are not defined.
MLEonly	If TRUE, then only the standard maximum-likelihood MNL model is estimated. No robust estimation model and no overdispersion parameter is estimated.

Details

The tanh estimator is a redescending M-estimator, and the LQD estimator is a generalized S-estimator. The LQD is used to estimate the scale of the overdispersion. Given that scale estimate, the tanh estimator is used to estimate the coefficient parameters of the linear predictors of the multinomial regression model.

If starting values are not supplied, they are computed using a multinomial multivariate-t model. The program also computes and reports nonrobust maximum likelihood estimates for the multinomial regression model, reporting sandwich estimates for the standard errors that are adjusted for a nonrobust estimate of the error dispersion.

Value

multinomRob returns a list of 15 objects. The returned objects are:

coefficients	The tanh coefficient estimates in matrix format. The matrix has one column for each formula specified in the model argument. The name of each column is the name used for the count variable in the corresponding formula. The label for each row of the matrix gives the names of the regressors to which the coefficient values in the row apply. The regressor names in each label are separated by a forward slash (/), and NA is used to denote that no regressor is associated with the corresponding value in the matrix. The value 0 is used in the matrix to fill in for values that do not correspond to a model formula regressor.
se	The tanh coefficient estimate standard errors in matrix format. The format and labelling used for the matrix is the same as is used for the coefficients. The standard errors are derived from the estimated asymptotic sandwich covariance estimate.
LQDsigma2	The LQD dispersion (variance) parameter estimate. This is the LQD estimate of the scale value, squared.
TANHsigma2	The tanh dispersion parameter estimate.
weights	The matrix of tanh weights for the orthogonalized residuals. The matrix has one row for each observation in the data and as many columns as there are formulas specified in the model argument. The first column of the matrix has names for the observations, and the remaining columns contain the weights. Each of the latter columns has a name derived from the name of one of the count variables named in the model argument. If count1 is the name of the count variable used in the first formula, then the second column in the matrix is named weights:count1, etc. If an observation has negative values specified for some outcome variables, indicating that those outcome alternatives are not available for that observation, then values of NA appear in the weights matrix for that observation, as many NA values as there are unavailable alternatives. The NA values will be the last values in the affected row of the weights matrix, regardless of which outcome alternatives were unavailable for the observation.
Hdiag	Weights used to fully studentize the orthogonalized residuals. The matrix has one row for each observation in the data and as many columns as there are formulas specified in the model argument. The first column of the matrix has names for the observations, and the remaining columns contain the weights. Each of the latter columns has a name derived from the name of one of the count variables named in the model argument. If count1 is the name of the count variable used in the first formula, then the second column in the matrix is named Hdiag:count1, etc. If an observation has negative values specified for some outcome variables, indicating that those outcome alternatives are not available for that observation, then values of 0 appear in the weights matrix for that observation, as many 0 values as there are unavailable alternatives. Values of 0 that are created for this reason will be the last values in the affected row of the weights matrix, regardless of which outcome alternatives were unavailable for the observation.
prob	The matrix of predicted probabilities for each category for each observation based on the tanh coefficient estimates.
residuals.rotate	Matrix of studentized residuals which have been made comparable by rotating each choice category to the first position. These residuals, unlike the student and standard residuals below, are no longer orthogonalized because of the rotation. These are the residuals displayed in Table 6 of the reference article.
residuals.student	Matrix of fully studentized orthogonalized residuals.
residuals.standard	Matrix of orthogonalized residuals, standardized by dividing by the overdispersion scale.
mnl	List of nonrobust maximum likelihood estimation results from function multinomMLE.
multinomT	List of multinomial multivariate-t estimation results from function multinomT.
genoud	List of LQD estimation results obtained by rgenoud optimization, from function genoudRob.
mtanh	List of tanh estimation results from function mGNtanh.
error	Exit error code, usually from function mGNtanh.
iter	Number of LQD-tanh iterations.

Author(s)

Walter R. Mebane, Jr., University of Michigan, [e-mail], http://www-personal.umich.edu/~wmebane

Jasjeet S. Sekhon, UC Berkeley, [e-mail], http://sekhon.berkeley.edu/

References

Walter R. Mebane, Jr. and Jasjeet Singh Sekhon. 2004. “Robust Estimation and Outlier Detection for Overdispersed Multinomial Models of Count Data.” American Journal of Political Science 48 (April): 391–410. http://sekhon.berkeley.edu/multinom.pdf

For additional documentation please visit http://sekhon.berkeley.edu/robust/.

Examples

# make some multinomial data
x1 <- rnorm(50);
x2 <- rnorm(50);
p1 <- exp(x1)/(1+exp(x1)+exp(x2));
p2 <- exp(x2)/(1+exp(x1)+exp(x2));
p3 <- 1 - (p1 + p2);
y <- matrix(0, 50, 3);
for (i in 1:50) {
  y[i,] <- rmultinomial(1000, c(p1[i], p2[i], p3[i]));
}

# perturb the first 5 observations
y[1:5,c(1,2,3)] <- y[1:5,c(3,1,2)];
y1 <- y[,1];
y2 <- y[,2];
y3 <- y[,3];

# put data into a dataframe
dtf <- data.frame(x1, x2, y1, y2, y3);

## Set parameters for Genoud
## Not run: 
## For production, use these kinds of parameters
zz.genoud.parms <- list( pop.size             = 1000,
                        wait.generations      = 10,
                        max.generations       = 100,
                        scale.domains         = 5,
                        print.level = 0
                        )

## End(Not run)

## For testing, we are setting the parmeters to run quickly. Don't use these for production
zz.genoud.parms <- list( pop.size             = 10,
                        wait.generations      = 1,
                        max.generations       = 1,
                        scale.domains         = 5,
                        print.level = 0
                        )

# estimate a model, with "y3" being the reference category
# true coefficient values are:  (Intercept) = 0, x = 1
# impose an equality constraint
# equality constraint:  coefficients of x1 and x2 are equal
mulrobE <- multinomRob(list(y1 ~ x1, y2 ~ x2, y3 ~ 0),
                      dtf,
                      equality = list(list(y1 ~ x1 + 0, y2 ~ x2 + 0)),
                      genoud.parms = zz.genoud.parms,
                      print.level = 3, iter=FALSE);
summary(mulrobE, weights=TRUE);

#Do only MLE estimation.  The following model is NOT identified if we
#try to estimate the overdispersed MNL.
dtf <- data.frame(y1=c(1,1),y2=c(2,1),y3=c(1,2),x=c(0,1))
summary(multinomRob(list(y1 ~ 0, y2 ~ x, y3 ~ x), data=dtf, MLEonly=TRUE))

Example output

Loading required package: rgenoud
##  rgenoud (Version 5.8-3.0, Build Date: 2019-01-22)
##  See http://sekhon.berkeley.edu/rgenoud for additional documentation.
##  Please cite software as:
##   Walter Mebane, Jr. and Jasjeet S. Sekhon. 2011.
##   ``Genetic Optimization Using Derivatives: The rgenoud package for R.''
##   Journal of Statistical Software, 42(11): 1-26. 
##

Loading required package: MASS
Loading required package: mvtnorm
## 
##  multinomRob (Version 1.8-6.1, Build Date: 2013/02/15)
##  See http://sekhon.berkeley.edu/robust for additional documentation.
##  Please cite as: Walter R. Mebane, Jr. and Jasjeet S. Sekhon. "Robust Estimation
##   and Outlier Detection for Overdispersed Multinomial Models of Count Data".
##   American Journal of Political Science, 48 (April): 391-410. 2004.
##


Equality constraints among parameters (after consolidation):
Equality constrained set 1 
outcome y1 regressor x1 
outcome y2 regressor x2 

Your Model (xvec):
                           y1 y2 y3
(Intercept)/(Intercept)/NA  1  1  0
x1/x2/NA                    2  2  0


multinomRob(): Grouped MNL Estimation
[1] "multinomMLE: hessian determinant: 973275091573.95"
[1] "multinomMLE: OPG determinant: 9175113261183712"
MNL LQD Fit: 2.903442 
MNL Estimates:
                                    y1         y2 y3
(Intercept)/(Intercept)/NA -0.07429061 -0.1130331  0
x1/x2/NA                    0.86051068  0.8605107  0

MNL SEs:
                                    y1          y2 y3
(Intercept)/(Intercept)/NA 0.011917062 0.011556630  0
x1/x2/NA                   0.008477768 0.008477768  0



multinomRob(): Calculating multinomial-t starting values.
Multinom-T LQD Fit (step  0 ): 1.238336 
Multinom-T Beta Estimates (step  0 ):
                                    y1          y2 y3
(Intercept)/(Intercept)/NA -0.01209041 0.000718553  0
x1/x2/NA                    1.00238612 1.002386121  0
Multinom-T Beta SEs (step  0 ):
                                    y1          y2 y3
(Intercept)/(Intercept)/NA 0.012473124 0.012371458  0
x1/x2/NA                   0.009725931 0.009725931  0
Multinom-T Omega Estimates (step  0 ):
            y1          y2
y1 0.004353550 0.001643915
y2 0.001643915 0.004403493
Multinom-T DF (step  0 ): 1.270218 

multinomRob(): Using multinomial-t estimates as starting values.
Multinom-T LQD Fit: 1.238336 
Multinom-T Estimates:
                                    y1          y2 y3
(Intercept)/(Intercept)/NA -0.01209041 0.000718553  0
x1/x2/NA                    1.00238612 1.002386121  0
Multinom-T DF: 1.270218 

multinomRob(): Starting Values 
                                    y1          y2 y3
(Intercept)/(Intercept)/NA -0.01209041 0.000718553  0
x1/x2/NA                    1.00238612 1.002386121  0
multinomRob(): starting fit = 1.238336 


LQD Results:
                                    y1          y2 y3
(Intercept)/(Intercept)/NA -0.01209041 0.000718553  0
x1/x2/NA                    1.00238612 1.002386121  0

LQD sigma: 1.238336 


(multinomTanh):
[1] "mGNtanh: number of Newton iterations 3"
[1] "mGNtanh: number of Newton iterations 1"
[1] "mGNtanh: hessian determinant: 589955666969.14"
[1] "mGNtanh: OPG determinant: 326208933849.713"
[1] "mGNtanh: tanh sigma^2: 0.884399205724957"
Tanh Estimates
                                    y1           y2 y3
(Intercept)/(Intercept)/NA -0.01092858 -0.001112076  0
x1/x2/NA                    0.99845470  0.998454702  0

Tanh Sandwich SEs
                                    y1          y2 y3
(Intercept)/(Intercept)/NA 0.010971825 0.011502659  0
x1/x2/NA                   0.008788404 0.008788404  0

TANH sigma: 0.940425 

Warning messages:
1: In fn(par, ...) : value out of range in 'lgamma'
2: In fn(par, ...) : value out of range in 'lgamma'
3: In fn(par, ...) : value out of range in 'lgamma'

Choice 1 : y1 Estimates and SE:
                Est SE.Sand t.val.Sand
(Intercept) -0.0109 0.01100     -0.996
x1           0.9980 0.00879    114.000


Choice 2 : y2 Estimates and SE:
                 Est SE.Sand t.val.Sand
(Intercept) -0.00111 0.01150    -0.0967
x2           0.99800 0.00879   114.0000


Choice 3 : y3 Estimates and SE:
            Est SE.Sand t.val.Sand
NA            0       0        NaN
NA            0       0        NaN


LQD sigma: 1.238336 
TANH sigma: 0.940425 

Number of Observations: 50 
Number of observations with at least one zero weight: 5 
Number of zero weights: 9 
TANH: weights
   name weights:y1 weights:y2
1     1   0.000000   0.000000
2     2   0.000000   0.899619
3     3   0.000000   0.000000
4     4   0.000000   0.000000
5     5   0.000000   0.000000
6     6   1.000000   1.000000
7     7   1.000000   1.000000
8     8   1.000000   1.000000
9     9   1.000000   1.000000
10   10   1.000000   1.000000
11   11   1.000000   1.000000
12   12   1.000000   1.000000
13   13   1.000000   1.000000
14   14   1.000000   1.000000
15   15   1.000000   1.000000
16   16   1.000000   1.000000
17   17   1.000000   1.000000
18   18   1.000000   1.000000
19   19   1.000000   1.000000
20   20   1.000000   1.000000
21   21   1.000000   1.000000
22   22   1.000000   1.000000
23   23   1.000000   1.000000
24   24   1.000000   1.000000
25   25   1.000000   1.000000
26   26   1.000000   1.000000
27   27   1.000000   1.000000
28   28   1.000000   1.000000
29   29   1.000000   1.000000
30   30   1.000000   0.590792
31   31   1.000000   1.000000
32   32   1.000000   1.000000
33   33   1.000000   1.000000
34   34   1.000000   1.000000
35   35   1.000000   1.000000
36   36   1.000000   1.000000
37   37   1.000000   1.000000
38   38   1.000000   1.000000
39   39   1.000000   1.000000
40   40   1.000000   1.000000
41   41   1.000000   1.000000
42   42   1.000000   1.000000
43   43   0.936756   1.000000
44   44   1.000000   1.000000
45   45   1.000000   1.000000
46   46   1.000000   1.000000
47   47   1.000000   1.000000
48   48   1.000000   1.000000
49   49   1.000000   1.000000
50   50   1.000000   1.000000

Choice 1 : y1 Estimates and SE:
            Est SE.Sand t.val.Sand
NA            0       0        NaN
NA            0       0        NaN


Choice 2 : y2 Estimates and SE:
               Est SE.Sand t.val.Sand
(Intercept)  0.693    1.22      0.566
x           -0.693    1.87     -0.371


Choice 3 : y3 Estimates and SE:
                 Est SE.Sand t.val.Sand
(Intercept) 2.60e-16    1.41   1.84e-16
x           6.93e-01    1.87   3.71e-01


Residual Deviance: 16.63553 

multinomRob documentation built on May 2, 2019, 6:54 a.m.