hw07 ready for grading

[heathersummers]

Please see here for hw07 folder.

[santiagodr]

Hi @heathersummers !!! here are my comments on your hw07

• your main README is beautifully organized and clear, as well as your README and folder for hw07. There are links to specific files and clear explanations (check on the Makefile link, that one doesn't work).

• Thanks for providing a very clear md. file, everything was easy to navigate from here.

• You downloaded the data in the Rmd file and also in the gapminder_exploratory.R file, which seems repetitive, but that's a minor point. This R file effectively reads the data, creates two plots, and compute a reordering of continent levels. This is also exported as a new file.

• The file gapminder_statistical.R, takes as input the file produced in the previous script. Run a set of linear models and extract parameters of these models in a new dataframe. And, also extract a subset of countries based on this new data and saved that table. You showed an excellent use of other resources learned in this course!.

• gapminder_figures.R takes again one input, and produce informative graphs for specific countries based on the information obtained from the previous analysis. The graphs are embedded correctly in the md. report as well as the tables.

• Also, there are plenty of clear and short comments in your Rscripts which is very helpful to understand what the code is going to do.

• Finally, the Makefile.R file seems to be correct, however I didn't use this strategy on my own homework, and I would guess that for me to be able to run the pipeline on this way, I would have to modify the folders to source the code?, not sure. Still, I learned from your hw how to use a makefile.R to write a analytical pipeline for several Rscripts!

• Overall, excellent assignment!!! All objectives are accomplished and you presented everything in a neat way, also providing interesting new graphs and data from gapminder!!!

[mattssca]

Hi @heathersummers and thanks for the invite!

Here is my peer review of your efforts for the second homework assignment in STAT547.

• I feel like you did a really good job with this task and there are defiantly some parts of your approach that I will implement in my own workflow.

• In my opinion you did everything that was required from you in this assignment. You demonstrate a thorough understanding of the material and your code is concise and appropriate comments are provided where needed.

• I really like your README.md file and how you've incorporated links that helps navigating your repository. All of your content in your repository is also very nicely organized.

• I like your reflections, and I can totally relate to what your saying about your initial struggle, I also had to go back to Jenny Bryans examples and they where very helpful.

• Besides some minor things like your Makefile link in your README.md file doesn't seem to work, I don't have much to comment on besides that i think you did an awesome job with this assignment.

I wish you the best of luck with all of your future assignments.

Cheers!

[ksedivyhaley]

Comment modified to reflect a more appropriate rubric for this assignment. Apologies for any confusion.

Three or more scripts, an Rmd, and a Makefile: OK (Makefile is R script) Starts by downloading data, ends with Rmd: OK (see comments) The output of each is the modified input of the previous step: No (figures does not input to statistical) Includes some analysis and at least one figure: Yes Makefile includes all scripts and Rmd with correct dependencies: No (is an R script) Makefile runs: No Bonus: Embedded figs in report Reflection: Yes

Comments:

• Your Makefile.R has directory problems.
• Your code runs from your current working directory, by default the same folder that your Rproj file is in, in this case STAT547-hw-Summers-Heather. Similarly, your scripts are saving output directly in the working directory. It looks like you've manually went and re-organized things into folders for figures, data, and scripts, but you should be coding this into the script.
• When writing or cleaning files, specify the subfolders to dig into, eg “hw07/Data files/gampinder.tsv” instead of just “gapminder.tsv”
• Your source() statements then run based on absolute directory names, which just plain fail when anyone tries to run it on a different computer (or if you go and change anything about the document path, eg changing the folder name of “STAT547-hw-Summers-Heather”). Again, you want a relative path: “hw07/R scripts/gapminder_exploratory.R."
• Also note: the spaces in the “Data files” and “R scripts ”folder names might cause you pain down the road.
• Once you fix the directories, the Makefile still does not run because you've used source() instead of rmarkdown::render() on your Rmd file.
• Your clean statement also removes your Rmd file, which is BAD because the Rmd is code, not output.

Other Comments:

• You should have one script, or one line in your Makefile, to download the data. Currently, you have it listed at the beginning of your gapminder_exploratory.R script, as well as in your Rmd file – which means you might end up redundantly downloading new data when all you want to do is update a plot.
• For exploratory data analysis, you have reordered your factors but then saved the data as a tsv – a format which doesn't preserve factor order. I'm also getting an error message that causes the plotting to fail: object 'country_colors' not found.
• Good use of nesting in your statistical analysis. I like that you also pull out the standard deviation.
• You've got magic numbers in your best/worst analysis, eg filter(std.error<0.0085|std.error>3.3)) : why are you filtering by those std error values?
• The link to your Makefile in your html report is broken.

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