search

Madhak1 / Dhakhwa-MADA-project

0 stars 2 forks source link

Part 3 Feedback - Cao #3

Open kcao1199 opened 7 months ago

kcao1199 commented 7 months ago

Hey Malika!

I really enjoyed reviewing your project and code.

  1. Data

Your data was easy to find and was very accessible. The raw-data folder clearly showed the csv file you initially procured, and the processed-data folder had the .rds file as expected from the processing .qmd.

  1. Processing

When reviewing your processing files, it ran without any problems. The files were properly saved, as indicated in the comment above.

  1. EDA

Similar to your processing file, your EDA .qmd and .R file both ran without any issues.

However, when I was rendered the .qmd, I saw that the table (gt table on line 101-123) did not form as expected for me. When comparing the 'printed' version to the png file you saved in the /product directory, it looks quite different from what I saw. I opened the eda_superstore.html file on my desktop and the same visual was shown on that file.

Aside from that, the .qmd and .R script both look great!

  1. Analysis

The following error was shown when attempting to run the analysis .R script. Due to that error, the top_model object was not able to be formed. 

Error in `show_best()`:
! `...` must be empty.
✖ Problematic argument:
• ..1 = "roc_auc"
ℹ Did you forget to name an argument?

I would recommend transferring this analysis to a quarto file for greater clarity.

Aside from that, the rest of your script ran fairly smoothly, and I really enjoyed reading the comments in your script. I particularly really liked how the script took the time to crunch the numbers and calculate those RFM scores. It felt like we were building a solid groundwork for our analysis, which makes me more confident about the results we'll get from the modeling part.

  1. Manuscript

I felt that your manuscript was very informative and well written. I particularly enjoyed the thoroughness of the data cleaning process and the detailed descriptive analysis, which provided valuable insights into customer behavior and sales trends.

  1. Overall

I noticed that the .R scripts and the .qmd varied quite a bit. Certain functions and analysis performed on the .R script was not shown on the quarto file and vice versa. Despite that, the README file suggested the quarto file will contain the exact same code as the R script or a quarto file pulling code from an R script. 

* First, there is an R script that you can run which does all the computations.
* Second, there is a Quarto file which contains exactly the same code as the R script.
* Third, my current favorite, is a Quarto file with an approach where the code is pulled in from the R script and run.

Overall, the project seems to be coming together. Aside from what I mentioned above, I would recommend updating a few of the README files to more accurately describe what your work is showing. I would also recommend either removing Dr. Handel's files or describing which files are yours in the README to ensure clarity for those who will be viewing your project.

Your project is very clear and easy to follow.

2/3

Madhak1 commented 7 months ago

Hey Kelly, Thank you so much for the thorough review and valuable feedback. I really appreciate your insights. I will look into the issues you pointed out and will make the best attempt to improve the project further.