NSKrstic commented 7 years ago

Peer Review:

Factor Management and Visualization Design:

Managed to filter out rows corresponding to Oceania, and dropped levels of the continent factor. But you should also remember to examine your other factors in your dataset (like country)! Their levels should be dropped too.
Unique and interesting choice to reorder the factor levels of country using the IQR for each country's life expectancy values
Covered what happens when arrange is used to change the order of values appearing in a figure. Found that only factor reordering has an effect.
I suggest being consistent with your colour legends between the figures, because it got kind of confusing to make the comparisons. Either that, or perhaps using a gradient between two different colours (a little more advanced, I highly recommend colorbrewer).

Saved a figure to file successfully. However, I suggest also looking at what happens if you don't specify the plot object within the ggsave function (it will save the plot last plotted, which might not be what you want).
Assessed whether read_csv and readRDS preserved factor reordering or not. I think you should emphasize this though.
Creative approach at recoding factor levels!
I see that you get a warning though when you're working on recoding factor levels. I believe this is because you didn't drop the factor levels of country? I may be mistaken.

I suggest having concise headers! Makes it an easier read.
Repository is mainly clean.
Remember to link to the homework in your issue.
Overall, well done and hits many of the key notes to take away from the assignment.

-Nikolas Krstic

sepkamal commented 7 years ago

Hi,

Nice work on hw5!

really like the comments in your code!
thanks for teaching me about droplevels. But I wonder why fct_drop didn't work. Did you make sure to just feed in $continent?
Did you think about dropping countries from country that are in Oceania?
Great job explaining the differences between the 3 plots
appreciate your use of head when showing the outcome of the .csv and .RDS files. Makes the assignment much easier to read, as opposed to having the whole table print out.
good job checking what's going in in your factors with levels and fct_count
I liked your idea to look at national sports!

I did a similar thing when importing data from the web. I cleaned up the file using excel to make it more manageable before importing into R.

Repo

I recommend giving your .md file a more informative name! This also applies to the list of hw's on your main readme file
otherwise repo is generally well organized!

Cheers, Sep

derekcho commented 7 years ago

Hi @navysealtf9k! Here are some comments about your hw05:

Your first task of dropping Oceania as a factor was correctly done. I would suggest showing the actual factor levels rather than just the counts so that the reader can verify that Oceania was dropped rather than reading through the code, however this is a minor point
Your second task was well done! I liked that you demonstrated more than one way to reorder factors using forcats
Your first plot in the exploration of arrange for plotting seems redundant as it is the same as the last plot except for a different color scheme. Having said that, I think that your plots were well done and illustrate the differences well. Filling by life expectancy was a nice touch as well since it allows the reader to see just how strongly life expectancy and gdp per capita are correlated.
Your explanations and demonstrations of the different read and write functions were well done!
Good work overall! Your report was well organized and you explained things well. In addition, you completed the additional task of adding an additional factor into the data as well. Good work in STAT 545A!

Note: your mark (check minus, check, check plus) will be distributed later.