My Guided Project: Exploring Ebay Car Sales Data

Hi,

Finished the Ebay Car Sales project. At this point, i’m overwhelmed with the range of methods() and arguments available to analyze the data. Definitely lot of practice is required to become well versed with the options. Kindly, provide the feedback.

ebay_project.ipynb (87.9 KB)

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

1 Like

Hi!

Congratulations on the project, I know every single one of them takes a lot of grit and googling to master.

Some of the things I notice that are different in my codes:

  1. I use inline comments above the code line, that way people read the comment first and the code second.
  2. I don’t use titles above code blocks. But it looks cool what you’re doing.

One thing which is very recognizable is the renaming columns with a dictionary. Right now I believe that’s a lot of typing. I mean it is personal preference. But as you print df.columns you can

  • copy that from the output
  • paste and change names
  • reassign back to df.columns.

Another thing you might want to correct is your end conclusion. Having high mileage is a bad thing for a car, which makes the value go down :wink:

One thing I noticed is the more technical the subject gets, the more important it becomes to know what every column means.

Finally, i would like to add that, you’ll get it. After hours and hours of googling I still don’t know the difference between .str.replace() and .replace() most of the time if one fails I try the other. Bottom line, try to find out what works best for yourself you’ll get there! Another thing which might come in handy is a cheat sheet. You can google most cheat sheets for every library :smiley:

Cheers, and keep on coding every day!

@DavidMiedema,

Thank you for the valuable feedback. Your tip to rename the columns is very helpful. I will implement it in future projects.

Regarding the conclusion, I forgot the point that less mileage is good for a car :slight_smile: . Now, I’ve reworked on the conclusion and have a different result. You can take a look at my new analysis if interested.

ebay_project.ipynb (102.5 KB)

I was not aware of this cheat sheets thing. So, I’m maintaining separate notes for quick reference. After googling I ended up in getting one from the blog itself - dataquest_pandas_cheat_sheet

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

@brudra
The most important thing is probably the ability to get the information you don’t know as fast as possible. Having some reference where you can find it is great.

Hence, apart from the cheat sheets and google, I have made a little folder of books e.g. the classic: Python Data Science Handbook by Jake van der Plas, or Think Bayes. Of course it is not done, but as a location independent person I hate physical books, so I download them.

Funny right! DQ made a cheat sheet but they never tell you explicitly in the course. I had to find out via google as well.

Your conclusion is better now! Mine was different though because I took 20 brands.
Have a good day :slight_smile:

1 Like