Description Usage Arguments Value See Also Examples
Based on an original dataset which includes a clustering variable, this
function creates a list of nboot
bootstrapped datasets with the
cluster variable (as opposed to rows) sampled with replacement. If using
complete case analysis, the function returns that list; if using multiple
imputation, the function returns a list of nboot
mice::mids
objects based on each of those bootstrapped datasets.
1 | create_bootdata(df, cluster_var, nboot, seed = NULL, impute = FALSE, ...)
|
df |
data.frame in "long" format. |
cluster_var |
character; matches one column name in |
nboot |
numeric; specifies how many datasets to create. Coerced to integer. |
seed |
numeric; if specified, sets seed for reproducibility. |
impute |
logical; whether to use |
... |
Additional arguments to pass to |
A list of length nboot
, where each element is either 1) a
data.frame
including all records from each sampled value of
df$cluster_var
, or 2) a mice
object based on one of the
bootstrapped datasets.
mice
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 | my_df <- data.frame(
id = sample(1:50, size = 500, replace = TRUE),
x = rnorm(n = 500),
y = rbinom(n = 500, size = 1, prob = 0.2)
)
## Without imputation:
create_bootdata(df = my_df, cluster_var = "id", nboot = 25)
## To demo imputation, make some data 10% MCAR
my_df_mcar <- my_df
my_df_mcar$x[sample(1:500, size = 50)] <- NA
my_df_mcar$y[sample(1:500, size = 50)] <- NA
## Use default arguments to mice()
create_bootdata(
df = my_df_mcar,
cluster_var = "id",
nboot = 25,
impute = TRUE
)
## Supply arguments to mice()
create_bootdata(
df = my_df_mcar,
cluster_var = "id",
nboot = 25,
impute = TRUE,
m = 10,
method = "mean"
)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.