think about how to make linear pool example more transparent

[elray1]

The code that is used to create the example for the linear pool, based on a mixture of three normal distributions, has been moved to a separate R script under the data-raw folder. On one hand, I understand the motivation for this. However, I also think this makes the example much less transparent -- what are these objects? how were they created? what is the format and content of the weights? what are the expected_quantiles, and why are we doing a test of equality in the example? In order to understand the example setup, I need to go look at help files for these objects, and this still doesn't answer some things (like the setup for the expected_quantiles), for which you really need to see the source code used to construct the example. And this code is not actually available anywhere in the package without going to the github repo source.

• My main idea/suggestion here is that we consider abandoning the use of this normal distribution example, and instead use example data from hubExamples, with the idea being that these standard example objects will already be familiar to package users (or at least, we hope they will become familiar after they see the objects used repeatedly, and we can point to documentation of the example data elsewhere).
• I would also advocate for thinking about a way to show the structure of the weights data as part of the example code -- either by just creating it within the example, or by showing head(weights).