stat545ubc-2021 / chloeyou_mini_data_analysis Goto Github PK

Hi Chloe,

Wonderful work on your first mini-data analysis deliverable. Please see detailed feedback below.

• Good start on your README.md, please add information about the project itself (at a high level) + a short description of each of the files in your repo + how to run code.
• Thanks for knitting your .Rmd file and uploading the knitted .md file to your repo. The knitted .md is well organized, nicely formatted, and overall easy to read!
• Quick note that you can use suppressMessages() around your library() call to avoid showing the messages when loading libraries.
• Ex 1.1: It would have been helpful to provide a short description/summary of each of these datasets rather than just list them.
• Ex 1.2: Great use of glimpse() function to explore the four datasets and a good summary of each dataset. You were also asked to "use your knowledge of dplyr to find out at least 3 attributes about each of these [four] datasets". So you needed to use two additional functions (e.g. class(), dim()) to explore data structure, variable types, etc.
• Ex 1.3: Good justification of your choice for the two datasets. It would have been helpful to use head() function to show what the actual data looks like.
• Ex 1.4: Excellent exploratory questions for each dataset.
• Ex 2.1: Excellent exploration of the steam_games dataset with a fantastic simultaneous use of both dplyr and ggplot2 functions.
  • Exercise 1: great idea to extract information on the ratio_of_postive_user_reviews and number_of_total_reviews, loved seeing the use of str_match() function. Since you created a new variable, you should have provided some information on it (e.g. a snapshot of values, distribution of values, some summary statistics).
  • Exercise 3: Fantastic graph, loved that you were able to find the share of missing values for each variable.
  • Exercise 4: Excellent and highly informative ridgeline plots - terrific! I would reorganize the categories in alphabetical order (with Action at the top, Strategy at the bottom).
• Ex 2.2: Excellent summaries of what you are trying to accomplish with each graph, helpful inline commenting, and great interpretation of the results of your analysis.
• Ex 3: Overall, excellent research questions that you should be able to answer in the next two deliverables.

Please let me know if you have any questions. Grades will be posted later today on Canvas.

Best,
Albina

Hi Chloe,

Overall, excellent work on your second MDA deliverable! Please see more detailed feedback below:

• Ex 1.1: Great research questions.
• Ex 1.2: Excellent summary tables and visualizations for all four questions, great analysis, and interpretation of the results.
  • Questions 1 (summary): you needed to print the output of each operation (in this case, print your summary table).
  • Question 2 and 4 (summary): this could have been re-written as case_when(ratio_of_postive_user_reviews < 0.25 ~ "negative", ratio_of_postive_user_reviews < 0.5~ "neutral", ratio_of_postive_user_reviews < 0.75 ~ "positive", TRUE ~ "very_positive").
  • Question 2 (graphing): the graph only has one geom_layer: geom_bar() - you could have added numeric values to the barcharts using geom_text()
• Ex 1.3: Excellent summary of what you are able to infer from the conducted analysis and what could be done next.
• Ex 2.1: Fantastic explanation as to why your data is tidy.
• Ex 2.2: Great transformation from tidy to untidy format and from untidy to tidy format.
• Ex 2.3: Great further analysis of the two selected questions, great use of various dplyr functions.
• Reproducibility, readability, and repo organization: Extremely well-organized repo with a detailed, informative, and clearly structured README.md. It's great that you added a README.md file in each of the milestone folders. For Milestone 3, make sure that each milestone's README.md file is a bit more informative - I'd add a couple of sentences about what you do in each deliverable. Milestone 2.Rmd file is well structured, nicely formatted, and easy to read. Thanks for knitting your .Rmd as .md document.

Please let me know if you have any questions. Grades will be posted soon on Canvas.

Best,
Albina

Hi Chloe,

Great work on your third MDA deliverable! Please see more detailed feedback below:

• Ex 1.1: Great illustration of fct_other() function to group multiple values into the "Other" category.
• Ex 1.2: Great illustration of year() function from lubridate package.
• Ex 2.1: Great use of lm() to build a multiple linear regression model. Your release_year should have been stored as numeric, not factor. You created both the train and the test datasets but you did not end up doing the testing?
• Ex 2.2: Good use of the tidy() function - you should have commented on the results.
• Ex 3.1: Great use of write_csv() and here::here() to save the summary table as a .csv document.
• Ex 3.2: Great use of saveRDS() here::here() to save the model statistics as a .RDS document.
• Tidy submission: Great README.md that orients one well in your repository. Well-organized repo with a great folder structure and knitted .md files for each milestone. Good use of releases to save different versions of your work.

Please let me know if you have any questions. Grades will be posted soon on Canvas.

Best,
Albina