Usage
missingdataexplorer(
data,
analysis_vars,
group_var,
time_var,
id_var,
pattern_analysis = TRUE,
mechanism_testing = TRUE,
correlation_analysis = TRUE,
temporal_analysis = FALSE,
group_comparison = FALSE,
mcar_test = "little",
min_pattern_freq = 0.05,
max_patterns_display = 20,
pattern_plot = TRUE,
correlation_plot = TRUE,
temporal_plot = FALSE,
upset_plot = TRUE,
cumulative_plot = FALSE,
monotonic_test = TRUE,
dropout_analysis = FALSE,
informative_missingness = TRUE,
completeness_threshold = 0.8,
case_completeness = TRUE,
variable_importance = TRUE,
chi_square_test = TRUE,
logistic_regression = FALSE,
survival_analysis = FALSE,
detailed_patterns = TRUE,
summary_statistics = TRUE,
clinical_interpretation = TRUE,
export_patterns = FALSE
)Arguments
- data
the data as a data frame
- analysis_vars
Variables to include in missing data pattern analysis
- group_var
Grouping variable for comparing missingness patterns (e.g., treatment arm, study site)
- time_var
Time variable for temporal missingness analysis (e.g., visit number, time since baseline)
- id_var
Subject identifier for longitudinal missingness analysis
- pattern_analysis
Analyze missing data patterns and frequencies
- mechanism_testing
Test missing data mechanisms (MCAR, MAR, MNAR)
- correlation_analysis
Analyze correlations between missingness indicators
- temporal_analysis
Analyze missingness patterns over time
- group_comparison
Compare missingness patterns between groups
- mcar_test
Method for testing Missing Completely At Random assumption
- min_pattern_freq
Minimum frequency for reporting missing data patterns (5\
max_patterns_displayMaximum number of missing data patterns to display in detail
pattern_plotGenerate missing data pattern plots
correlation_plotGenerate missingness correlation heatmap
temporal_plotGenerate temporal missingness plots
upset_plotGenerate UpSet plot for pattern visualization
cumulative_plotGenerate cumulative missingness over variables plot
monotonic_testTest whether missingness follows monotonic patterns
dropout_analysisAnalyze dropout patterns in longitudinal data
informative_missingnessTest whether missingness is informative (related to outcomes)
completeness_thresholdThreshold for acceptable variable completeness (80\case_completenessAnalyze completeness at the case/subject levelvariable_importanceAssess which variables predict missingness in otherschi_square_testChi-square tests for independence of missingnesslogistic_regressionLogistic regression to predict missingness patternssurvival_analysisSurvival analysis for time to dropout/missingnessdetailed_patternsShow detailed information for each missing data patternsummary_statisticsInclude comprehensive summary statisticsclinical_interpretationProvide clinical interpretation and recommendationsexport_patternsExport missing data patterns for external analysis A results object containing:
Comprehensive missing data pattern exploration and visualization for clinical research. Analyzes missing data mechanisms, patterns, and correlations to inform imputation strategies. Includes advanced diagnostics for missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR) assessments. Essential for understanding missingness before imputation and for regulatory documentation of missing data handling strategies. data('clinical_data')missingdataexplorer( data = clinical_data, analysis_vars = c("primary_endpoint", "biomarker", "demographics"), group_var = "treatment_arm", pattern_analysis = TRUE, mechanism_testing = TRUE )