Remove variables that fail a missing data test from models

Comparing missingness by looking at a table is good but we also want to update models to exclude missing data from the predictors.

Usage

remove_missing(
  df,
  ...,
  label_fn = label_extractor(df),
  significance_limit = 0.05,
  missingness_limit = 0.1
)

Arguments

df

a dataframe of individual observations. If using the tidyselect syntax data grouping defines the intervention group and should be present. if the formula interface is used the first variable in the right hand side of the formula is used as the intervention, in which case grouping is ignored.

...

a list of formulae that specify the models that we want to check

label_fn

(optional) a function for mapping a co-variate column name to printable label. This is by default a no-operation and the output table will contain the dataframe column names as labels. A simple alternative would be some form of dplyr::case_when lookup, or a string function such as stringr::str_to_sentence. (N.b. this function must be vectorised). Any value provided here will be overridden by the options("tableone.labeller" = my_label_fn) which allows global setting of the labeller.

significance_limit

the limit at which we reject the hypothesis that the data is missing at random.

missingness_limit

the limit at which too much data is missing to include the predictor.

Value

a list of formulae with missing parameters removed

Examples

df = iris %>%
  dplyr::mutate(Petal.Width = ifelse(
    stats::runif(dplyr::n()) < dplyr::case_when(
      Species == "setosa" ~ 0.2,
      Species == "virginica" ~ 0.1,
      TRUE~0
    ),
    NA,
    Petal.Width
  ))
remove_missing(df, ~ Species + Petal.Width + Sepal.Width, ~ Species + Petal.Length + Sepal.Length)
#> subtype_count summary for Petal.Width
#> subtype_count summary for Sepal.Width
#> subtype_count summary for Petal.Length
#> subtype_count summary for Sepal.Length
#> fisher test on Petal.Width
#> fisher test on Sepal.Width
#> fisher test on Petal.Length
#> fisher test on Sepal.Length
#> More than 10% of data is missing for variables Petal.Width.
#> Data is missing not at random (compared to Species) at a p-value<0.013 (0.05 over 4 comparisons) for variables Petal.Width.
#> [[1]]
#> . ~ Species + Sepal.Width
#> <environment: 0x55eda93213b0>
#> 
#> [[2]]
#> . ~ Species + Petal.Length + Sepal.Length
#> <environment: 0x55eda93213b0>
#>