Missing Value Imputation for Accelerometer Data
Missing Value Imputation for Accelerometer Data
Daily Activity Plot
accelmissing
Create a Missing Flag Matrix
Imputation by PMM under ZIP model.
Imputation by PMM under ZIPLN model.
Imputation by Bayesian ZIPLN model.
Computing Missing Rate
Select the Valid Days
Include or Exclude Subjects by Criteria
Proportion of Wearing over Time
Imputation for the missing count values in accelerometer data. The methodology includes both parametric and semi-parametric multiple imputations under the zero-inflated Poisson lognormal model. This package also provides multiple functions to pre-process the accelerometer data previous to the missing data imputation. These includes detecting wearing and non-wearing time, selecting valid days and subjects, and creating plots.