hw02 ready for grading

[sepkamal]

Hi @vincenzocoia @gvdr @ksedivyhaley @joeybernhardt @mynamedaike @pgonzaleze @derekcho

Please see my hw02 folder: https://github.com/sepkamal/STAT545-hw-Kamal-Sepehr/tree/master/Hw02

Thanks, Sep

[ksedivyhaley]

Smell test of data: Yes Explores at least one categorical/quantitative variable: Yes Uses various ggplot types: Yes (dot, histogram) Uses filter(), select(), and %>%: Yes Bonus (evaluate code, tables, more dplyr): Yes: evaluate code, plus arrange() and mutate() Reflection on process: Yes

Comments:

• typeof() produces odd results with factors – it's calling $country and $continent integers because they are stored as integers under the hood, but it's important to be able to distinguish that from actual numeric data like $lifeExp. The 'class()' function will tell you when something is a factor.
• Rule of thumb: start a pipe with the dataset (eg gapminder)
• “This shows that for each of the 12 entries present for each country, each one corresponds to a different year (with 5 year intervals). ”: technically you've only proven that Canada works this way. If you wanted to be sure that all countries used the same 12 years, what could you do?
• Good Top 10 table – uses the functions in a strategic & cohesive way
• Nice job on the text angle for your dot graph!
• Note that arrange() doesn't affect the display of the graph, just the table.
• Nice job with the line on the mean for your histogram.
• It's a bit hard to tell where the label for Canada is pointing in your last plot – but it's a very neat feature to use!
• I think there's a more elegant way to define rel_pop than making a separate vector for Canadian population. Also, didn't think that mutate should require you to define your new variable ahead of time. Would have to do some playing around to confirm.

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

[cheungamanda]

Hi Sep, Nicely done! Here's my review for homework 02:

• Your repo is easily accessible! I would recommend creating another folder for your homework 01 for more organization.

• Great job using headers and text to guide the reader along. I would suggest listing the questions to explain what you're doing.

• Your typeof(gapminder$continent) code says continent is an integer, when it's a factor, but you also answered that it was factor. Perhaps explain discrepancy in your code when necessary!

• Good work exploring individual variables using tables and figures!

• I like your use of various plot types! Your plot of the top 30 European countries vs. population is very nicely done. I like that you incorporated piping, filter(), and head(). I also learned how to angle text from your code, so thank you! I'm going to incorporate that in my future plots.

• I didn't know about the geom_vline() function, and I liked that you used it to mark the mean! Great idea (I'm going to try this too)!

• I like that you used different features and I learned how to individually label data points from your Canada label. It's hard to see which point it's referring to, but definitely useful for other data sets!

• I would consider adding more to your reflection and maybe adding links to the sources you found helpful (I appreciate that final image)! Overall, great work!

Thanks, Amanda

[CassKon]

Accomplishments

• Make a homework repo – Made a repo with hw02 as a folder and created a readme file with link to homework
• Bring rectangular data in – loaded data and packages for assignment
• Smell test the data – answered all questions asked in this part of assignment using functions in R
• Explore individual variables  - Explored continent and lifeExp
• Explore various plot types –Scatterplot, frequency polygon, histogram, Boxplot, Area chart,
• Use filter(), select() and %>% - used all of these and used piping to feed data into ggplot
• But I want to do more! – answered the filter problem and added correct function  

  Overall comments:

• Used heading to separate parts of assignment
• Experimented with different plot types and customized plots within
• I like the comments throughout, especially commenting on whether the graph is a good or bad representation of the data. Also the picture at the end is great.
• It shows that you explored the customization of different plots such as labeling points and adding different layer types (geom_vline)
• Changing the axis labels would increase interpretation of plots
• more reporting on your process (details!) would be one improvement for next time, as well as pointing out resources used.

Well done, plot on!

-Cassandra