Multiple Imputation with Denoising Autoencoders


[Up] [Top]

Documentation for package ‘rMIDAS’ version 0.4.1

Help Pages

add_bin_labels Reverse numeric conversion of binary vector
add_missingness Apply MAR missingness to data
coalesce_one_hot Coalesce one-hot encoding back to a single variable
col_minmax Scale numeric vector between 0 and 1
combine Estimate and combine regression models from multiply-imputed data
complete Impute missing values using imputation model
convert Pre-process data for Midas imputation
import_midas Instantiate Midas class
midas_setup Manually set up Python connection
mid_py_setup Configure python for MIDAS imputation
na_to_nan Replace NA missing values with NaN
overimpute Perform overimputation diagnostic test
python_configured Check whether Python is capable of executing example code
python_init Initialise connection to Python
set_python_env Manually select python binary
skip_if_no_numpy Skip test where 'numpy' not available.
train Train an imputation model using Midas
undo_minmax Reverse minmax scaling of numeric vector