Investigate ways to support data

There are few places where data really becomes important. Tests, notebooks, and generally for the package.

We've seen many instances where a package requires a large amount of data to run. But it's impractical to include that data in the distribution. We've come up a few specific ways to address that, but let's consider if there's a way that we can include a best practice approach in the template as well.

首先审查 template 当前如何处理测试、notebook 和包数据，然后评估正在考虑的数据管理方法。完成的标准是选择并记录一种可以纳入 template 的最佳实践方法。