hw09 ready for grading

[mylinhthibodeau]

Dear colleagues,

I had some last minute troubleshooting challenges, so I hope the package is at least functional at this stage. It's not that I didn't try, believe me, but I was unable to resolve some ongoing issues, and if the package is not functional, please notify me !!

• powers3 repository here
• README.md here
• Vignette Rmd here

Thank you for your time and consideration,

Regards, My Linh Thibodeau

[CassKon]

Homework 9

• Really great job on the package, I can tell that you put a lot of effort into it
• Documentation is thorough and license + description files are complete (with exception of the package documentation that you already noted)
• Test files are completed with at least different inputs in each one
• Namespace file is in order and vignette was made for package
• Definitely covered making three functions and in fact went above and beyond providing three categories of functions
• Detailed readme and comprehensive notes on package - I like that you provided the code needed to install your package and I was able to install the package just fine using that code!
• All sources listed and an inspirational quote at the end - GREAT!

Amazing job once again and I find little room for improvement beyond the problems that you identified. I was a little confused by the superpowers function and I got the message:

Error in file(file, "rt") : cannot open the connection In addition: Warning message: In file(file, "rt") : cannot open file 'R/data/superpowers.csv': No such file or directory

This could be something that I am doing wrong on my part however! Overall it was a pleasure to read/review!

[ilgan]

@ mylinhthibodeau

Functions:

• I won't describe all the functions, they all work just fine, the description showed up for each one of them, just as it should.
• your functions showed good understanding on how and why the errors have to be dealt with. I also saw it through your readme file.
• plus for the right license

Data:

• in csv format. I am pretty sure that is teh reason I cannot run it on my machine. You have to convert it to an .rda file so it wil be usable with the package. It will still run on your PC since you do have this file locally. Installing package won't download the csv (just saw that @CassKon got the same issue).

Readme:

• the readme file is informative.
• thanks for explaining how you actually worked on the projects and what were your problems/bugs + the solutions.

Overall the nice work!

[ksedivyhaley]

At least one new function: Yes Function argument defaults: Yes Assertions for validity checks: Partial (missing and needed in find_factors, see comment) Three unit tests per function (success/failure): OK (see comments) Document exported functions: Partial (see comments) Passes check(): Yes Updated README and vignette: Partial (vignette not properly rendered) Package in new repo, reflection in old: Link in old, detailed reflection in new

Comments:

• You shouldn't be including arguments in your package documentation, just your function documentation.
• “"Since you provided character data instead of numeric data, I took the liberty of using the sum of all characters in your vector (which adds up to 5), as the input numeric value for the function.” : I literally winced when I found this. Fail quickly and fail loudly! If you're trying to cube non-numeric data it's likely an error. You do not want to do something clever to work around it, ESPECIALLY not without (1) stating it in the documentation and (2) throwing an actual warning so that downstream coders know something fishy is going on – printing a message isn't enough since you can miss those if not executing code manually in small chunks. (In fact, some people think that generating a formal warning isn't enough and you should just fail and throw an error.)
• pow does the same thing and includes an example that demonstrates the function, but it needs to be more obvious because this is DANGEROUS.
• pow(3, 5, plot_it=TRUE): “I know you wanted a plot, but it's not really worth it to plot only one data point!” Message is OK but it would still be nice to produce a plot (or again at least mention this behavior in the documentation).
• Generally, you seem to do a lot of guessing about what the user wants or doesn't want when they provide unexpected input. Some of what you do is less dangerous than other stuff but it is usually better to fail with an informative error message if something looks fishy.
• “return("This function only accepts integers!")” should be “stop("This function only accepts integers!")” so we get an error message and not an output that unexpectedly but contains a string instead of a list of integers.
• It would be good to state that find_factors returns NULL if there are no factors.
• Note: Booleans should probably throw an error with find_factors and new_factorial but they don't because they're treated as 0/1 here
• largest_prime and find_factors say that they return vectors (value) but actually return lists.
• The superpower() function could also use more description of the structure of its output (which is not just a list, but specifically a data frame).
• Because cube and pow don't fail where they should, your tests also lack a test for failure – they instead test that the function avoids failure in the way you want it to.

Your mark will be distributed later. If you would like more feedback, please feel free to message me on slack.