Clean and Analyze Employee Exit Surveys by Enrique Martinez

Share Guided Project
#1

Hello, I finished my project, thanks a lot to @artur.sannikov96 , to @Raghav_A, and to @ivelinagenova, I saw your excellent projects and gave me a lot of fresh ideas of what to do, I tried to explore other result.

Clean and analyze employee exit survey 3.ipynb (2.8 MB)

Click here to view the jupyter notebook file in a new tab

8 Likes
Announcing the Community Champions for this week!
#2

Hey @jemartinezm1

Firstly Congratulations for winning the Community Champions of this Week! :tada:

Your entire Project is really cool and I liked all the visuals which speaks the content :slight_smile: One thing I noticed is few heatmaps are truncated and I think it’s good to set heatmaps “Limits” manually so that your visuals would be more crystal clear.

Thanks again for sharing this with community.

Best
K!

2 Likes
#3

Thank you @prasadkalyan05 for your kind words, I’m going to check the heatmap issue, I think that it is not very useful when you have a lot of data, so the next time maybe I use it only with a small sample.

1 Like
#4

That’s a great idea! Good luck!

Announcing the Community Champions for this week!
#5

Hello @jemartinezm1! Thanks for sharing. You did a great job, your project is well-organized, has clear objectives and conclusions.

Here is some feedback:

  • No need to use %matplotlib inline twice. Once activated, this magic command works everywhere in your project.
  • Great job on understanding the columns with missing data (about aborigens and NESB).
  • You do not seem visualize all missing values in the TAFE survey (first 0:31 and then 36:71) - it’s in case of the matrix.
  • Don’t write this: There is not clear criteria for this part, because is the first time that I’m doing this in python I’m following the instructions of the project.! You have to be sure in what you are doing even if your not sure.
  • When you drop the columns why do you do that? Could you think of a motivation? It’s ok to to write your thought and “I don’t know” when you’re doing the project but when you share it with others it’s better to avoid them (especially if you want this project as a part of your portfolio).
  • It’s better (for readability) to import all libraries in the first code cell.
  • You can logically order your categories (like from New to Veteran) both in tables and plots.
  • You had an interesing idea and solution color your tables according to the number in each cell! I added it to my table of curios code solutions:)
  • The plot “Dissatisfied employees according to their position” is not very clear:
    • It’s very long
    • The categories are not in a logical order
    • The tick labels are very small
    • Are the data labels essential for each bar?
    • Generally try to understand what is the most important information you want to communicate and create a plot based on it.
  • Are you sure about the teachers? These results may be a fruit of a bigger sample of teachers in the dataset. Try to see if you have any bias.

That’s all for me.

Happy coding:)

1 Like
#6

Arthur excellent review, thank you so much for your generosity, I’m going to take note of your specific observations, I’m glad you learn something new with the code. Happy coding!

2 Likes