Steps Before Modeling: Cleaning to EDA

Before being able to utilize the machine learning algorithms we all know and love, it is important to understand how to initially get the data in proper order and how to effectively analyze the data in order to produce the best model. This post will provide a basic roadmap that a data scientist can follow to perform exploratory data analysis (EDA) when working with a new dataset in Python. For this post, we will examine SAT and ACT participation rates by state for 2017 and 2018. The datasets include statewide average scores for each exam section, composite/total scores for both exams and participation rates. After loading the file into a Pandas dataframe, it is useful to look at the first few rows and the size of the dataset to get a feel for what you are working with.