detect.miss.MNAR.MAR: detect.miss.MNAR.MAR It detects if the missigness depends on...

In BeanLabASU/metabimpute: Missingness Evaluation, Simulation and Imputation in Metabolomics

detect.miss.MNAR.MAR It detects if the missigness depends on other varibales Correlates an indicator matrix (0 not missing , 1 missing) with the original data that can help determine if variables tend to be missing together (MAR) or not (MCAR). Kabacoff, Robert I. R in Action. manning, 2010. From juuussi/impute-metabo

Description

detect.miss.MNAR.MAR It detects if the missigness depends on other varibales Correlates an indicator matrix (0 not missing , 1 missing) with the original data that can help determine if variables tend to be missing together (MAR) or not (MCAR). Kabacoff, Robert I. R in Action. manning, 2010. From juuussi/impute-metabo

Usage

detect.miss.MNAR.MAR(data, alpha = 0.05, correlation_method)

Arguments

data	, matrix with missing values
alpha,	significance level , default is 0.05

Value

results list of vectors : MissingVar = vector containing the columns numbers of the data matrix with missing values that are less than 80 PairsCorVar = data frame containing the pairs of correlated variables in the data matrix, MAR_MNAR = vector containing the columns numbers of the data matrix with MAR or MNAR missingness ExcludedVar = vector containing the columns numbers of the data matrix with missing values that are more than 80

Examples

marietta <-detect.miss.MNAR.MAR (miss_data)

BeanLabASU/metabimpute documentation built on Feb. 5, 2023, 11:41 p.m.