Giters

moderndive / ModernDive_book

Statistical Inference via Data Science: A ModernDive into R and the Tidyverse

Home Page:https://www.moderndive.com/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Potential improvements for v2

rudeboybert opened this issue · comments

commented

@avaldivi6

  • Erase Bowl from sections
  • Footnote for almonds
  • Make sure activity can be replicated by instructors (e.g. went to chemistry department to borrow scales). Be clear this refers to instructors (e.g, "If you are an instructor...")

General

  • Add Arturo's bio and photo to Preface.
  • Think about moving regression chapters all to the end Keeping where they are to ensure some exploratory regression is done before inferential.
  • Currently in v1 CLT is a subsection of Section 7.5. For v2, consider moving it to its own section to make CLT more prominent. Can do this by rebasing #326 to eventual v2 branch that will get created.
  • Refactor Section 7.3 to break down reading into more bite-sized pieces. Rebase #324
  • Per Katie’s suggestion, add discussion on R^2 and R^2-adjusted in Subsection 6.3.1 on Model selection. In particular somewhere after "visual model selection" we do. Can use moderndive::get_regression_summaries() wrapper to broom::augment() to get at these values easily. Can reuse fc4621d. Also per Will’s suggestion, add visualizations illustrating the breakdown of variances that go into R^2.
  • Add recap/key concepts to each chapter AND/OR define as learning goals at the outset of each chapter
  • In Ch7 Conclusion, consider adding a section on random sampling vs assignment?
  • Add exercises at the end of each chapter
  • Extend {infer} to more advanced use cases using fit() function
  • Use base-pipe |> instead of %>%
  • Remove any outstanding HTML comments. <!-- --> (They can be really dangerous if not closed!)
  • Address the warning message explicitly for group_by() in text and fix index.Rmd to remove options(dplyr.summarise.inform = FALSE)
  • Add relocate() to end of Chapter 3
  • Update preface with updates to 2nd edition
  • Add almonds data to moderndive package
  • Ask Kelly for an updated foreword
  • Create branch and link to first edition. 2nd edition stays at moderndive.com
  • Change font size of PDF to one point smaller
  • Add diagrams for sampling and bootstrapping

nycflights23

  • Use the anyflights package to create an nycflights23 package to replace nycflights13 data.
  • Update the early_january_weather and alaska_flights data frames in moderndive package to use '23 data
  • Update Learning Check solutions in Chapters 1-4 for '23 data
  • Explain that nycflights23 is an updated version of nycflights13 using the anyflights package
commented

In Section 9.6.1, the discussion about the warning message to check that conditions have been met for the theoretical method {infer} might be rewritten for clarity. The warning message is generated by a visualize() command with method = "both". However, the discussion about the warning message follows a get_p_value() command using a simulation-based null distribution (null_distribution_movies_t), where this warning doesn't apply (right?). It might be clearer to discuss the warning message before getting the p-value. Or get_p_value might be run instead with a null distribution generated using assume(distribution = "t"), where the warning message would then apply.