domdf / dce_guidance

DomDF

A personal R package containing functions for Bayesian data analysis and plotting. These include:

• gen_cov_exp_quad( ) A function for generating an n-dimensional exponenitated quadratic covariance matrix (in various formats)
• tidy_mcmc_draws( ) A function for extracting draws from a CmdStanR object (fitted Stan model) in long, tidy format
• VoPI( ) A function for a value of perfect information analysis, for a generic decision making under uncertainty problem
• ...as well as some webscraping and plotting functions.

This can be installed using the following code:

require(devtools)
devtools::install_github(repo = 'DomDF/DomDF')

dce_guidance's Issues

Adding some BLUF + other comments

Some thoughts on the initial draft in mid-November:

1. Adding Bottom Line up Front

• Stronger introduction sentences that capture the significance of the work, such as:

New data streams and collection methods offer opportunities for significant improvements to safety and efficiency of the build environment. But how do engineers decide how much data to collect, or the quality and granularity required for the data they do collect in order to make an informed decision? In other words, how do they quantify the value of the information they must gather?

• Thoughts on this addition to the title: Value of Information in Data-Centric Engineering?
• Thoughts on adding the Key Concepts section earlier so readers can identify what parts are most relevant to them?

2. Adding conclusion & next steps

• This is not necessarily the purpose of this document, but it could be helpful to conclude with some suggestions for what to read next, how they can use the tools you've created, and how they can contribute to this field.

3. Linking real world examples

• Would it be appropriate to link to some of the related LR/Turing case studies that are alluded to?

4. Other

• Thoughts on grounding the Example Calculations sections in key questions that each section aims to answer?