mice {mice}R Documentation

Multivariate Imputation by Chained Equations

Description

Produces an object of class "mids", which stands for 'multiply imputed data set'.

Usage

mice(data, m = 5, 
    imputationMethod = vector("character",length=ncol(data)), 
    predictorMatrix = (1 - diag(1, ncol(data))),
    visitSequence = (1:ncol(data))[apply(is.na(data),2,any)], 
    defaultImputationMethod=c("pmm","logreg","polyreg"),
    maxit = 5, 
    diagnostics = TRUE, 
    printFlag = TRUE,
    seed = NA)

Arguments

data A data frame or a matrix containing the incomplete data. Missing values are coded as NA's.
m Number of multiple imputations. If omitted, m=5 is used.
imputationMethod Can be either a string, or a vector of strings with length ncol(data), specifying the elementary imputation method to be used for each column in data. If specified as a single string, the given method will be used for all columns. The default imputation method (when no argument is specified) depends on the measurement level of the target column and are specified by the defaultImputationMethod argument. Columns that need not be imputed have method "". See details for more inromation.
predictorMatrix A square matrix of size ncol(data) containing 0/1 data specifying the set of predictors to be used for each target column. Rows correspond to target variables (i.e. variables to be imputed), in the sequence as they appear in data. A value of '1' means that the column variable is used as a predictor for the target variable (in the rows). The diagonal of predictorMatrix must be zero. The default for predictorMatrix is that all other columns are used as predictors (sometimes called massive imputation).
visitSequence A vector of integers of arbitrary length, specifying the column indices of the visiting sequence. The visiting sequence is the column order that is used to impute the data during one iteration of the algorithm. A column may be visited more than once. All incomplete columns that are used as predictors should be visited, or else the function will stop with an error. The default sequence 1:ncol(data) implies that columns are imputed from left to right.
defaultImputationMethod A vector of three strings containing the default imputation methods for numerical columns, factor columns with 2 levels, and factor columns with more than two levels, respectively. If nothing is specified, the following defaults will be used: pmm, predictive mean matching (numeric data); logreg, logistic regression imputation (binary data, factor with 2 levels); polyreg, polytomous regression imputation categorical data (factor >= 2 levels).
maxit A scalar giving the number of iterations. The default is 5.
diagnostics A Boolean flag. If TRUE, diagnostic information will be appended to the value of the function. If FALSE, only the imputed data are saved. The default is TRUE.
printFlag
seed An integer between 0 and 1000 that is used by the set.seed function for offsetting the random number generator. Default is to leave the random number generator alone.

Details

Generates multiple imputations for incomplete multivariate data by Gibbs Sampling. Missing data can occur anywhere in the data. The algorithm imputes an incomplete column (the target column) by generating oappropriate imputation values given other columns in the data. Each incomplete column must act as a target column, and has its own specific set of predictors. The default predictor set consists of all other columns in the data. For predictors that are incomplete themselves, the most recently generated imputations are used to complete the predictors prior to imputation of the target column.

A separate univariate imputation model can be specified for each column. The default imputation method depends on the measurement level of the target column. In addition to these, several other methods are provided. Users may also write their own imputation functions, and call these from within the algorithm.

In some cases, an imputation model may need transformed data in addition to the original data (e.g. log or quadratic transforms). In order to maintain consistency among different transformations of the same data, the function has a special built-in method using the ~ mechanism. This method can be used to ensure that a data transform always depends on the most recently generated imputations in the untransformed (active) column.

The data may contain categorical variables that are used in a regressions on other variables. The algorithm creates dummy variables for the categories of these variables, and imputes these from the corresponding categorical variable.

Built-in imputation methods are:

norm
Bayesian linear regression (Numeric)
pmm
Predictive mean matching (Numeric)
mean
Unconditional mean imputation (Numeric)
logreg
Logistic regression (2 categories)
logreg2
Logistic regression (direct minimization)(2 categories)
polyreg
Polytomous logistic regression (>= 2 categories)
lda
Linear discriminant analysis (>= 2 categories)
sample
Random sample from the observed values (Any)
Special method
If the first character of the elementary method is a ~, then the string is interpreted as the formula argument in a call to model.frame(formula, data[!r[,j],]). This provides a simple mechanism for specifying a large variety of dependencies among the variables. For example transformed versions of imputed variables, recodes, interactions, sum scores, and so on, that may themselves be needed in other parts of the algoritm, can be specified in this way. Note that the ~ mechanism works only on those entries which have missing values in the target column. The user should make sure that the combined observed and imputed parts of the target column make sense. One easy way to create consistency is by coding all entries in the target as NA, but for large data sets, this could be inefficient. Moreover, this will not work in S-Plus 4.5. Though not strictly needed, it is often useful to specify visitSequence such that the column that is imputed by the ~ mechanism is visited each time after one of its predictors was visited. In that way, deterministic relation between columns will always be synchronized.

For example, for the j'th column, the impute.norm function that implements the Bayesian linear regression method can be called by specifying the string "norm" as the j'th entry in the vector of strings.

The user can write his or her own imputation function, say impute.myfunc, and call it for all columns by specifying imputationMethod="myfunc", or for specific columns by specifying imputationMethod=c("norm","myfunc",...).

side effects: Some elementary imputation method require access to the nnet or MASS libraries of Venables & Ripley. Where needed, these libraries will be attached.

Value

An object of class mids, which stands for 'multiply imputed data set'. For a description of the object, see the documentation on mids.

Author(s)

Stef van Buuren, Karin Oudshoorn, 2000

References

Van Buuren, S. and Oudshoorn, C.G.M.. (1999). Flexible multivariate imputation by MICE. Report PG/VGZ/99.054, TNO Prevention and Health, Leiden.

Van Buuren, S. & Oudshoorn, C.G.M. (2000). Multivariate Imputation by Chained Equations: MICE V1.0 User's manual. Report PG/VGZ/00.038, TNO Prevention and Health, Leiden.

Van Buuren, S., Boshuizen, H.C. and Knook, D.L. (1999). Multiple imputation of missing blood pressure covariates in survival analysis. Statistics in Medicine, 18, 681-694.

Brand, J.P.L. (1999). Development, implementation and evaluation of multiple imputation strategies for the statistical analysis of incomplete data sets. Dissertation, TNO Prevention and Health, Leiden and Erasmus University, Rotterdam.

See Also

complete, mids, lm.mids, set.seed

Examples

data(nhanes)
imp <- mice(nhanes)     # do default multiple imputation on a numeric matrix
imp
imp$imputations$bmi     # and list the actual imputations 
complete(imp)       # show the first completed data matrix
lm.mids(chl~age+bmi+hyp, imp)   # repeated linear regression on imputed data

data(nhanes2)
mice(nhanes2,im=c("sample","pmm","logreg","norm")) # imputation on mixed data with a different method per column

[Package mice version 1.15 Index]