Member-only storySoner YıldırımFollowTowards Data Science--ShareThe web is a highly-valuable asset as a data source. For instance, a substantial amount of training data used to create large language models comes from the web.However, it’s usually not in the most suitable format. Web data is mainly unstructured (i.e. in the form of free text). Even if it has a predefined structure, web data requires lots of cleaning and preprocessing before it can be used for analytical purposes.In this article, we’ll take a messy Dataset that includes the price and some other attributes of cars and clean it using the pandas library.You can download the dataset from my datasets repository if you want to follow along and execute the code yourself. It’s called “mock_car_dataset”. Some of the operations we’ll perform on this messy dataset are as follows:I created the dataset with mock data. However, it’s just like a car dataset you’d scrape from the web. I know it because I’ve done it before.The dataset is in CSV format. Let’s start with creating a pandas DataFrame from this file.The dataset contains 20 rows and 6 columns, which means we have data of 6 attributes for 20 cars. Although it’s a small dataset, the operations we’ll do can easily be applied to much larger datasets (i.e. hundreds of thousands of rows)Let’s see what these attributes are ( cars.head() will show you the following ):----Towards Data ScienceData Scientist | Top 10 Writer in AI and Data Science | linkedin.com/in/soneryildirim/ | twitter.com/snr14Soner YıldırıminTowards Data Science--3Khouloud El AlamiinTowards Data Science--23Soner YıldırıminTowards Data Science--8SuryaCreatX--RaviTeja GinLevel Up Coding--Chitranjan GuptainCodeX--Avi ChawlainTowards Dev--5Ann Mary ShajuinTowards AI--2Thomas Kidu--1HelpStatusAboutCareersBlogPrivacyTermsText to speechTeams