Final Project for Martin, Katherine and Zach (original location w branching: https://github.com/zelson71/final)
Title The Search for Value in Wines in the US and France and the Relative Price / Rating Relationship Using Data Made Available by Wine Enthusiast Magazine.
Project Goals
-
Determine if Californian wine represents a better value than French wine using price data and wine ratings.
-
Predict wine price with region / subregion and rating data.
Specifically, compare price versus rating for both Chardonnay and Cabernet Sauvignon / Bordeaux Red Wine Blends.
Answer the following questions: Is there a clear value winner? Does rating level change the answer? Is it possible to predict wine prices from these reviews? What is the impact of region and subregion on pricing and rating? From a wine review perspective, what words are more likely to be used as rating increase?
Prerequisites Requirements to install and implement: Scikit Learn, Tensorflow: Keras, Matplotlib, SQLalchemy, Pandas, Postgres, Tableau (Professional or Student version)
Installing Open and run "Cleanup and Database creation.ipynb"
- Locate this file in "Database_Creation"
- Edit code under heading "Create A SQL Connection and Define the Database" where the connection to Postgres is established and personalize - DO NOT change database name.
Executing Open and run "US vs French Wine Analysis.twb"
- Locate this file in "Tableau"
- For a high-level view of the analysis refer to "US vs French Wine Overview.pptx"
Authors Zach Elson, Katherine Sullivan and Martin Wehrli