search

aumath-advancedr2019 / Sampling

Package for making Permutation and Bootstrapping in R
https://aumath-advancedr2019.github.io/Sampling/
0 stars 0 forks source link

Peer review (2) #2

Open Hoejrup opened 4 years ago

Hoejrup commented 4 years ago

1. Purpose of the package

The Description file does a good job describing the package purpose and function. I did find a small typo: “The bootstrap will return a #blot of the sampling distribution of a given estimate, it's standard error and confidence intervals. “ The vignette gives nice and detailed information about the functions implemented in the package and is nicely structured. I like the added menu at the top. This is a nice touch, that I will need to implement in our package :-).

2. Completeness

It’s nice that you have added data into the package. This makes it easy to test and see how the package can be useful. Also easy to run the examples. Nice that you’ve added examples in the documentation.

3. Code quality and sophistication

The installation ran smoothly after I installed a few missing packages. The code is really nicely structured and you make great use of a different sophisticated method such as S3-Object-oriented programming and implemented nice warnings, that makes it easy to see if you make “stupid” choices of parameters. I like that you have added the choice for “selfchosenmethod” or “my_method”, this makes this even more useful, for “super-users”.

4. Documentation

The functions are well documented, combined with the vignette. The Vignette, provided very useful documentation, on how to use the different S3-methods for different objects, and also how the different output is stored as classes. Also contains detailed explanations of output and examples. I would maybe like, an explanation of the math behind it, f.x. how you compute the p-value.

5. Interface

The interface feels natural and in line with what I’m used to from other packages. You’ve implemented nice warnings, that makes sense, both to the user but also for the functionality of the package. Also nice with the addition of ggplot, if installed, else we use the built-in plot function.

6. Conclusion

It’s easy to see, that there has been put in a lot of effort, to make the package stream-lined and professional. You have used a lot of the material presented in this course, like warnings, S3-object-oriented methods, ggplot and many more. I found a few typos (nothing affecting the functionality of the documentation), that you might want to have a look at. Also, I would really like to see a bit more math explained, f.x. formulas for p-value. This would make it even easier to use, since you don’t have to check the source code, to understand the “mechanics” of the functions. In the end, a nice package.

LotteHind commented 4 years ago

Thank you for your feedback. You mention that you needed to install some missing packages before the functions could run, what packages is that?

Hoejrup commented 4 years ago

To be honest, I can't remember 😅. I just installed them without much thought.