Reinigen Sie die Autos!

Hello all, please see my latest project titled:

Reinigen Sie die Autos!!!
An Analysis of ebay Car Sales in Germany

Exploring_eBay_Car_Sales.ipynb (392.9 KB)

Sometimes I get a little lost in the weeds and I haven’t completed a hypothesis test in several years, so let’s hope it makes a bit of sense!

Please enjoy the weekend!

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

2 Likes

Hi @shaun.oilund,

Thanks for sharing your as-always thorough and well-structured project. It’s a pleasure to read.

To be honest, I am quite low on energy right now, so I apologize that I can’t provide thorough comments which your project deserves.

Anyhow, here are some of my thoughts:

  1. I’m a bit unsure about my thoughts on the extensive use of text bolding. It’s important yet at the same it seems that the bolded words are too frequent. It’s also a matter of style and I don’t have a reference in hand that clarifies how bolding should be used (and how frequent). I wonder if italics might be a better fit for some of the bolded words or maybe no text style at all.
  2. Cell [1] and [2] can be merged without losing anything. Maybe this is another style thing because opinions clash a bit on when imports should be written when it comes to Jupyter.
  3. Another stylistic alternative: for the variables and their descriptions, you can also use Markdown table. It puts them all in neat left-aligned cells which is easier to read.
  4. I really like that you added the modified raw data set and intermediate/cleaned data set (it’s like a checkpoint of sorts and I find that pretty neat; some people might want to skip the whole cleaning and only want the cleaned data sets).
  5. Some of the code comments can be redundant or unnecessary especially if you’ve mentioned what you’re trying to do in the text explanation and if the method names are self evident e.g. .info is clearly about displaying info about the data frame so a better comment might be what info you’re specifically looking for instead of the general “info”. Then again, this is context dependent. Most of the time I reckon, those who read the codes (say someone in this forum) will likely be somewhat literate in Python and pandas so they should at least know the basic methods and patterns; if you’re writing for (very fresh) beginners e.g. tutorials, then yeah, being explicit and clear on every single methods and functions can be useful.
  6. Maybe use a more layman-ish title for the plots e.g. modify the “autos[‘price_EUR’] Price Distribution” and “autos_price_EUR_2022[‘price_EUR_2022’] Price Distribution”. Plots are a good place to reduce the cognitive load and speed up the conveying of information because of their visual nature; using code-like titles introduce a little bit of load even for programmers. Other than that, non-technical readers who don’t understand indexing will probably scratch their heads. Other than that, maybe also add the year to the first plot title (or any titles that don’t reference the year) to make it clear that it’s for the 2016 price.

Awesome work and keep it up. Cheers.

3 Likes

Hi @wanzulfikri,

Thank you very much for taking the time to view and comment on my project. Your comments are always thorough and detailed as well which is a big help; it’s very much appreciated! I was feeling a little under the weather too, I hope you are feeling better and can find a jolt of energy to get you back to true form!

You’re right about the text bolding, I did get more than a bit carried away with all that and definitely will tone it down. There probably is a standard out there somewhere, but I’m not aware of one.

As for cells [1] and [2], I see what you mean, merging the cells I think would clean up that section.

I never thought about using a Markdown table like that, I’ll try it out and see how it looks.

The code comments are redundant, it boils down to knowing your audience as you mentioned and having a balance between adding clarity or clutter; adding code comments to every cell leans towards clutter for sure.

I’m not sure why I gave my plots titles like that especially when I could have just used the titles in the table of figures, code brain I suppose!

Thanks again for your comments, they do help a lot, especially when you have the horse blinders on!

Take care, relax, and be well!

Shaun

2 Likes

No worries @shaun.oilund.

Haha I like that term - “code brain”. Yeah, it happens to all of us and it’s always important to remind ourselves that codes are not a natural way of expressing thing even if it feels (somewhat) natural like that for us programmers. It happens a lot when we’re too immersed in our code and afterwards forget to put on the non-technical hat when writing the story for the data.

Anyhow, I hope you’ll also feel better soon. Cheers.

2 Likes