I’m currently checking your project out as I’m writing this. The points I’d like to state are:
If you upload this project to your github etc. as a folder you downloaded from Dataquest containing the .csv file and the notebook, I’d highly recommend that instead of linking the second dataset to Dataquest, saying that the dataset is in the repository. Because not everyone might have access to that Dataquest mission page, i.e non-Dataquest people.
I like how you’re excited about this project in the markdowns! I also like the fact that you’re commenting the code. Commenting is a really good habit!
I’d recommend using autos.head() instead of printing the entire dataset in the second cell. And since you’ve used that in the cell below, you can omit the second cell.
In the 5th cell, I see that you’ve replaced the column names one-by-one in the function you defined. Instead of that, I’d recommend you use the map() function.
“In a way it could be tempting to throw away the rows with a price of 0, considering it ‘incorrect data’. But then, is a price of 1 more realistic? Or a price of 29, 65 or 123? It is hard to set a particular limit for this, and consider everything below that ‘outliers’ or ‘incorrect’. And maybe people do just want to get rid of their old worn-out cars? Available for who still sees some value in it - maybe just for spare parts - and wants to put the effort to collect the cars.” is an interesting approach. However, spamming in these kinds of websites considered, I’d still recommend you put a lower limit for the price column. If you don’t do that, summary statistics calculations like the mean price etc. might come out very wrong and unrepresentative of the actual data.
The beginning of the markdown below the 29th cell “Nothing looking particularly suspect there actually.” should be changed to “Nothing looking particularly suspicious* there actually.”.
Don’t be discouraged although I’ve made so many recommendations, that’s because I looked into your project in great care. Overall, great work! Keep the good work up and keep sharing with us!
Once more: thank you for your feedback!! (As communicated to you in a private message a couple of months ago, I really appreciated your feedback, but had no chance to respond to it back then. I had to pause my study for multiple months, and just recently I restarted it. For the record, for what it’s worth still: let me still respond.)
Makes perfect sense. Unfortunately, github is not something that I am using yet, not very familiar with it yet. It seems that I will learn about this later during my Dataquest learning.
Fully agree. Learning: do not just follow the instructions (as this was a literal instruction actually to just type auto), but make it your ‘own’ project.
Fair point. I argued that any particular value being set as the lower limit would be a rather arbitrary one. However, setting at least a value as lower limit then is probably a better solution than not setting any at all.