Project 1: Electric Vehicles
Electric Vehicle Clean Alternative Fuel Vehicle (CAFV) Elegibility
Problem Statement
I looked at an interesting dataset pertaining to Electric Vehicles with their model, make, state, and CAFV Eligibility. I wondered to myself, "Which EV are popular in certain cities in the US?". So, I decided to base my problem statement on what EV incentives should certain American cities implement depending on what type of EV is popular.
Dataset
I found this dataset from kaggle.com that is about electric vehicle (EV) populations which includes various EV models with just under 224,000 records and 17 columns within the dataset. Some of these columns include:
- Vehicle Identification Number (VIN)
- County
- City
- State
- Model
- Make
- Make
- Electric Vehicle Type (Hyrbrid or Fully Electric)
Pre-processing the Data
Firstly, I eliminated the unnecessary columns that weren't part of location statistics or CAFV Eligibility.
Then, I checked to see if there were any records with null values and I found three records that were null for county and city and dropped them from the dataset.
Lastly, I decided to see how many unique Cities, counties and states there were in the dataset and found there were too many cities for me to use, so I dropped the City column from the dataset.
Data Understanding and Visualization
To start off, I created a scatterplot with the x-axis being each of the 45 listed states, the y-axis being each EV brand, and the hue being the CAFV Eligibility Status (Eligible, Not Eligible, not enough research to tell)
It seemed like there was a lot of records on the state of Washington which includes every EV/hybrid brand on the market. It's very surprising to see so many Tesla models where CAFV Eligibility is unknown. This could be due to a brand new Tesla model or a lack of incentive from most of the states to research Tesla since Tesla is the largest EV manufacturer. There is also a lot of unknown dots in the state of Washington which could be due to Washington's numerous records in the dataset or something else. The only EV brand that had CAFV Eligiblity across all the states that it was in was Chrysler. The last metric to point out is that there are a lot of unknowns for Rivian, Volkswagen, Subaru, and Hyundai.
I then made two scatterplots of each EV brand that are in roughly 5 or more states. The first scatterplot had each model as the x-axis and the states at the y-axis whereas the second scatterplot had the electric vehicle type (hybrid or fully electric) on the x-axis and the states on the y-axis. These scatterplots were made to see which models and vehicle type from each brand had CAFV Eligibility or not.
*Only the Tesla scatterplot is shown here. For the rest of the scatterplots, click on the link that says "My Jupyter Notebook" at the bottom of this cell
Summary
The general trend is that most EV brands -- with the exception of Tesla -- have unknown research for CAFV Eligiblity for their Battery Electric Vehicles (BEV) compared to their hybrid counterparts which mostly have a defined CAFV Eligibility based on their battery range. This trend is consistent with most states that have these brands, which means the issue of EVs not having CAFV Eligibility is not based on location, but on make and vehicle type. This leaves a lot of untapped potential for some EV brands having CAFV Eligibility for their cars, allowing more people to buy EVs.
What is the impact of this project?
We can say from this dataset that Fully Electric Vehicles need to be researched more to qualify for Clean Alternative Fuel Vehicle (CAFV) Eligibility so that more people from every state can afford and purchase more electric vehicles, allowing the electric vehicle industry to grow. This is especially true for certain car brands that aren't selling many EVs and want more sales.
Reference
EmptyAd. (2025, February 7). Electric Vehicle Population Data. Kaggle. https://www.kaggle.com/datasets/avis02/electric-vehicle-population-data