Is it possible to specify the width of the histogram when stratified sampling a numerical sequence?

Machine Learning and Modeling
,
1

Is it possible to specify the width of the histogram when stratified sampling a numerical sequence?

Use the following link to learn.

Stratified sampling of numerical data is divided into four parts.
Presumably, this is a quartile range.

Is it possible to specify more than four partitions if I want to sample the data more precisely to make the model learn better?

thank you!

2

Yes, there is the breaks parameter just for that.

> args(rsample::initial_split)
function (data, prop = 3/4, strata = NULL, breaks = 4, pool = 0.1, 
    ...)