Nothing
count_NA()
counts the number of missing values in a vector, data frame or
matrix.apply_imputation()
and friends now have an option to convert tibbles
instead of throwing an error.delete_MAR_1_to_x()
and delete_MNAR_1_to_x()
now handle unordered factors
as documented (thanks to Steve Roehrig for reporting).delete_MAR_1_to_x()
and delete_MNAR_1_to_x()
now display the correct
adjusted x
value, if it is too high or too low (thanks to Steve Roehrig
for reporting).apply_imputation()
type rowwise now works for data frames (thanks to
@khughitt for fixing).delete_values()
now only takes mech_type
and derives mechanism
.delete_
functions have the argument n_mis_stochastic
now. For some
functions, this is only a renaming of the old stochastic
argument (e.g.
delete_MCAR()
), for others this is completely new. The new name emphasis
that this argument controls if the number of missing values is stochastic
or deterministic. delete_MAR_1_to_x()
and delete_MNAR_1_to_x()
get a new argument
x_stochastic
along the line of n_mis_stochastic
.missMethods.warn.too.high.p
to control the
displaying of warnings for too high values of p
(the probability for a
value to be missing).delete_values()
and get_NA_indices()
centralize many steps of the old
(not exported) delete_
functions. delete_MAR_
and delete_MNAR_
functions and delete_MCAR()
call the
new delete_values()
function now. delete_
functions use the new get_NA_indices()
to determine
the missing values. impute_EM()
now returns the number of performed EM iterations as attribute.delete_rank()
now hands the argument ties.method
over to rank()
.delete_one_group()
(wrong argument FUN
instead of
cutoff_fun
).median.factor()
(thanks to @labachevskij).impute_LS_adaptive()
has now the default setting warn_r_max = FALSE
.impute_in_classes()
allows to apply any imputation function inside imputation classes impute_hot_deck_in_classes()
hot deck imputation inside of imputation classes (adjustment cells)impute_EM()
imputes values using EM parameter estimatesimputed_expected_values()
imputes expected values from a multivariate normal distributionimpute_LS_adaptive()
performs LSimpute_adaptive as described by Bo et al. (2004)impute_LS_array()
performs LSimpute_array as described by Bo et al. (2004)impute_LS_combined()
performs LSimpute_combined as described by Bo et al. (2004)impute_LS_gene()
performs LSimpute_gene as described by Bo et al. (2004)cov_only
and cor_only
as parameter
in evaluate_imputation_parameters()
cols
variables: now all should be named cols_mis
, cols_ctrl
etc.ds
variables: now all should be named ds_imp
, ds_orig
etc.pars
variables: now all should be named pars_est
or pars_true
cols_seq
is now correct, if the donor is only one numeric valueFunctions for the creation of missing values:
delete_MAR_censoring()
and delete_MNAR_censoring()
create missing (not) at random values using a censoring mechanismdelete_MAR_one_group()
and delete_MNAR_one_group()
create missing (not) at random values by deleting values in one of two groupsdelete_MAR_rank()
and delete_MNAR_rank()
create missing (not) at random values using a ranking mechanismFunctions for evaluation:
evaluate_imputation_parameters()
compares estimated parameters after imputation to true parametersdelete_MAR_1_to_x()
and delete_MNAR_1_to_x()
can now handle (unordered) factorsevaluate_imputed_values()
and evaluate_parameters()
: six forms of NRMSE, nr_equal, nr_NA and precisionevaluate_imputed_values()
: add argument cols_which
to select columns for evaluation. delete_
functions now take the same first three arguments: ds
, p
, cols_mis
Functions for the creation of missing values:
delete_MCAR()
creates missing completely at random values in different waysdelete_MAR_1_to_x()
and delete_MNAR_1_to_x()
create missing (not) at random values using a 1:x mechanismFunctions for imputation:
impute_mean()
, impute_median()
, impute_mode()
different forms of mean, median and mode imputationimpute_sRHD()
simple Random Hot-Deck imputation with the possibility to specify a donor limitapply_imputation()
a function to apply aggregating functions for imputationFunctions for evaluation:
evaluate_imputed_values()
compares imputed to true valuesevaluate_parameters()
compares estimated to true parametersMiscellaneous:
median.factor()
computes medians for ordered factorsAny scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.