A Unique Star Wars Survey Project 2

Please replace the previous file “Star Wars Survey2” with this one “Star Wars Survey3”. I just learned how to properly embed a picture. Thank you kindly. All feedback is welcome.

https://app.dataquest.io/m/201/guided-project%3A-star-wars-survey/9/next-steps

Star Wars Survey3.ipynb (763.8 KB)

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

2 Likes

Hi Bruce,

Wow, you are making really very fast progress in learning from one project to another, congratulations indeed! :+1: I’m afraid soon I will not be able to suggest you anything new, because everything will become just perfect! :sweat_smile: It’s clearly visible that you are a fan of Star Wars, since you were really enjoying conducting this analysis. Perfect overall project structure, interesting and capturing storytelling, all the necessary subheadings, observations, links, markdown emphasizing: fonts, numbered lists, but what’s more - even colors! It was a good idea to divide long code lines into several lines for better readability. Awsome and very clean visualizations, with a good choice of colors. Also, I liked the analysis by age category, and introducing Kruskal-Wallis Non-Parametric Test was something completely fresh and incredible! Great job!

Some comments from my side, hopefully still useful :grinning:

  • It’s better to keep code comments as laconic as possible. All the explanations and observations go to markdown. Also, you might consider removing the empty line between a code comment and the code itself. Everything will be still perfectly readable, only that it helps to save the vertical space of the project.
  • If the neighbor code cells don’t have any output or markdow explanations between them, it’s better to combine them into one code cell, even if they contain different informational blocks (like [3] and [4] or [21] and [22]).
  • I’d suggest you to remove those backslashes from question numbers, they are really scaring :slightly_smiling_face:.
  • For some bar plots, you can consider rotation of x-tick labels ([5], [7], [9]), otherwise now they are vertical. The code cell [9]: better to remove spines also here.
  • For the bar plots with a lot of bars ([22]), a good idea is to create a horizontal plot instead of vertical.
  • The code cell [6]: here you might consider using a function to optimize the code.
  • In the code cells with a lot of print() statements (for example, [2]), it’s better to insert an empty line between different outputs (print('\n')).

I hope my ideas were interesting. Amazing job your project, keep up this level!

1 Like

Thank you Elena! Great suggestions for improvement.
You will always have great ideas for improvement.
You are well ahead of me in knowledge, understanding, coding and writing skills.
Blessings on your day! Bruce

One other thing. It looks like the wrong file was discarded.
I requested the file with number 2 at the end to be discarded and the one with number 3 to be kept.
Maybe I should resubmit and remove all the back slashes.
Bruce

Hi Bruce,

Thatàs cool that my suggestions were helpful! Yes-yes, I deleted exactly the file with 2 and reviewed the file with 3. The cover picture is perfectly dispalyed in this new file, by the way.

Thank you Elena! Have a wonderful day!
Bruce

1 Like

Thank you, Bruce, you too! Good luck with your future projects! :relaxed:

Elena:
I tried to use Community Q & A to pose a couple of questions, but the format required me to input code that I had trouble with. My questions have nothing to do with a code problem, so I’m using this mail system if it’s ok with you.

Question 1:

I’ve learned that GITHub does not take Jupyter type files (*.ipynb). What file format must I convert my Guided Project files to and what is the simplest safest way to do it?

Question 2:

So far, on many occasions during my learning, I’ve come across pip packages (e.g. squarify, … etc.) and other non-pip packages (wordcloud, … etc.) that are valuable. What is the simplest and safest way to download these? I’ve read that downloading any of these from command line can be dangerous if not done properly. Also, where can I get a list of all the available packages/libraries that can be used for Python/Jupyter …?

Thank you kindly Elena.
Bruce

1 Like

Hi Bruce,

As for posting questions, just ignore (or delete) the section where they ask you to put the code and write your question in text, like here. This format is not obligatory there.

Q1. Why, github actually takes ipynb files. I always downloaded them (and only this type) in my github without any problem :slightly_smiling_face: So just download your ipynb files without any preliminary conversion.

Q2. I would say just the opposite: it’s better to install (or deinstall) all the libraries through pip, and not through conda. They are all installable from pip, and it’s much quicker and safer: in this way, you will not have several versions of the same library, which can have conflicts between them. For installing wordcloud, use pip install wordcloud.

Thank you Elena!
So, do I execute the “pip install wordcloud” command from the MS DOS environment, or from ??

Elena, I’m using Windows PowerShell to execute pip install commands.
I did successfully install “squarify”.
I consistently get this error statement for wordcloud no matter what combination of install coding used.
“Running setup.py install for wordcloud … error”

I did update python to 3.9 and pip to 21.0.1
Bruce

Bruce, about wordcloud, you should run pip install wordcloud in Anaconda Prompt (also for any other installations, in future). Then, already from your Jupyter notebook in Anaconda, you should run import wordcloud. It should work smoothly, please try and let me know. I did the same when installing and importing it. And with squarify everything is ok?

Thank you again Elena.
It was successful.
I did download 30 other python modules which included squarify using Windows PowerShell.
Bruce

1 Like