Package edu.cmu.tetrad.data.missing
package edu.cmu.tetrad.data.missing
-
ClassDescriptionEstimates the covariance matrix of data in which some values are censored rather than missing: a value at a ceiling (a grade of 100, a detector at its maximum, a score at the top of its scale) is known only to be at least that bound, and a value at a floor to be at most it.The search wrapper implementing
MissingDataPolicy.MULTIPLE_IMPUTATION: impute m completed datasets, run the given algorithm on each, and pool the resulting graphs by edge frequency using the existingGraphSamplingmachinery (so pooled graphs carry per-edge probabilities and render with the same GUI visualization as bootstrap results).The result of a multiple-imputation search: the pooled graph (with per-edge frequencies attached by GraphSampling) and the individual per-imputation graphs.A "lite" chained-equations (MICE-style) multiple imputer for continuous, discrete, or mixed data, using predictive mean matching (PMM) as the single imputation engine for both variable types.A descriptive audit of the missingness in a dataset, intended to be computed (and logged) before a search is run on real data, and to give the user the evidence needed to choose aMissingDataPolicy.The result of Little's MCAR test: the chi-square statistic, its degrees of freedom, the p-value, and pattern accounting.An explicit statement of how missing values in a dataset are to be handled by a score, independence test, or search wrapper.An immutable specification of how missing data are to be handled: aMissingDataPolicytogether with the parameters that policy needs (EM settings forMissingDataPolicy.EM_COVARIANCE, the number of imputations forMissingDataPolicy.MULTIPLE_IMPUTATION, and so on).How the effective sample size used by penalized scores (e.g., the n in BIC) is determined when the analysis runs on a covariance matrix estimated from incomplete data.Small shared utilities for the missing-data policies: listwise deletion, resolution of a possibly-nullMissingDataSpecto a policy (with a logged warning where the historical behavior was a silent switch), a cheap missingness summary suitable for logging from constructors, and the effective-sample-size calculation for analyses run on covariance matrices estimated from incomplete data.Works out what dropping every variable whose missingness rate exceeds a threshold would do to a dataset, without dropping anything.A variable that a threshold would drop, with what is needed to judge whether dropping it is safe.The consequence of applying one threshold.A capability declaration for scores and independence tests, stating what the component itself can do with data containing missing values.An interface for multiple-imputation methods: given a dataset with missing values, produce m completed copies in which every missing entry has been replaced by a draw from an imputation model, with observed entries unchanged.A multiple imputer for continuous data under a saturated multivariate normal model.Estimates a covariance matrix from data with missing values by pairwise deletion, without a model of the data.Rubin's rules for pooling a scalar estimate across the m completed data sets of a multiple imputation.A pooled parameter.Test-wise (family-wise) covariance estimation for components whose local calculations consume a covariance submatrix and a sample size: for a requested column set, the sample covariance matrix is computed over exactly the rows that are complete on those columns, together with the number of such rows.The covariance matrix of a column set over the rows complete on that set, the number of such rows, and the rows themselves.Computes, and caches, the row subsets used by test-wise deletion: for a given set of columns, the rows of a dataset on which all of those columns are observed.