i think that the vocabulary is mixed in the docs between “collections” and “arrays”. for example, in a datatable, the datatypes of a datasource includes “array”, when what it really takes is a collection.
Arrays are a good choice when you have a simple dataset that you need to iterate over quickly. It’s generally faster and use less memory, especially for small datasets.
Collections are generally slower and use more memory, especially for large datasets. However, they provide more functionality and flexibility. You can use it when you need to work with more complex datasets that require filtering, sorting, or transformation.