ACM Research coding challenge (Fall 2022)

CHRISTOPHER SHEPPARD

This semester's challenge is especially open-ended. Here is a dataset on Kaggle called "CarsForSale". It contains data scraped from the online car marketplace Cars.com. Each row contains 25 pieces of information about a car's listing, such as its price, year, model, and color.

The challenge is to do something interesting with the data. Can you find a pattern, answer a question, or create a visualization?

3 step proccess

1. Clean Data

2. Find Patterns in Data

3. Visualize Useful information

1.1 DATA

Dispay the data types and columns

1.2 PRICE TO INT

Convert Price string to an int for future processing

RESULT: Price is now a float 64

1.3 Used/New

Convert all certifications to certified

RESULT: Now only Used and Certified are the only unique values

2.1 ConsumerRating

Find topmost variables correlated with ConsumerRating RESULT: CustomerRating Mean is 4.702762961382547

RESULT: These 4 columns have a high correlation to consumer rating ValueForMoneyRating 0.917873,

ReliabilityRating 0.914597,

ComfortRating 0.860040,

PerformanceRating 0.805849,

2.2 All CORRELATIONS

Display all correlations

2.3 PRICE CORRELATIONS

Display all price correlations

RESULT: Price has some correlations with Year and an anti-correlated to Mileage

2.4 MAKE

Find most common car sold

3.1 HISTOGRAMS

Display all availible in histograms

3.2 VALUE TYPES

Display final data types

3.3 HOT ENCODING

Initiate Pandas Hot encoding for future predictive modeling