Star Wars Survey : Your feedbacks are welcome

Hi colleagues,
Your feedback will be highly appreciated.
Thank you.

Star Wars Survey.ipynb


Hi @sabritrabelsi90! Thanks for sharing.

Overall, I liked your project, it has a great logical structure, analysis and well-done conclusions.

You can add the questions you want to answer in the introduction, so we will understand what is interesting early in the project.

Also consider improving your pie plots (it’s even better if you change them to another type) since it’s quite difficult to understand what you are trying to say with them.

Happy coding!

1 Like

Hi @sabritrabelsi90, in my opinion is a nice proyect. Very extensive and with detail. Good job!

Because you use many times value_counts() Series method, may be it could be usefull for you: if you use the apply() function with a subset of your dataframe in combination with a lambda function which use value counts. I mean, in you have a subset(a few columns to which you want apply value counts), you cand do so: subset.apply(lambda col: col.value_counts()). This manner you apply value counts to all the columns of your subset. I hope if could be useful for your.

Thanks for sharing!


1 Like

Hi @artur.sannikov96! Thank you for the feedback.
I am totally agree with you, adding the questions we want to answer in the introduction is a good practice so i made the necessary modifications, i have referenced the cells that are linked to the questions.
Regarding the pie plots could you tell me please what kind of improvement do you suggest ?


1 Like

Hi @Daniel_H, Thank you for the feedback.
I think that what you have suggested is convenient for columns that share the same unique values otherwise the returned DataFrame will have a lot of Nan values. isnt?

Yes, you’re right. For this proyect I think it’s useful, because there are some columns with the same unique values (e.g. unknown columns, fan and seen question, etc…). For columns with different unique values, you can also use describe() method for a subset, and the include = “all” as argument. Although this method won’t give you all the unique values, it can serve as preliminary analysis. Regards.
Daniel. Vamos!

Hi @sabritrabelsi90! I hadn’t received a notification, sorry about that.

You could transform your pie charts into bar charts that are clearer and easier to understand.

Hi @sabritrabelsi90

Project looks good! Also can tell you are a fan after all you answered the question of did Han shoot first :smiley:

1 Like

@sabritrabelsi90 I liked your charts, you managed very well to present your analysis +1 for pie charts and overall conclusion

1 Like