Runs queries on imdb dataset.

cli.py is the main script for driving the application. Run python cli.py help for general usage.

To load data into a database, run python cli.py load-data.

To rank movie genres by average profitability, run python cli.py rank-genre [num_genres].

To rank actors/directors by average profitablity, run python cli.py rank-personnel [num_persons].

The application is structured using a simple MVC pattern.

In the model layer, data is represented using SQLAlchemy as the ORM and backed with sqllite3 by default.

In the controller layer, function arguments are passed as key-value arguments and abstracted from the view or data source.

In the view layer, arguments and data are passed from a user or client, parsed, invokes controller actions, and returns formatted data.

The data/ directory holds input data, logs, and the database itself.

Database connection string, defaults to local sqlite file in data.

Path name to input data set, defaults to "data/movie_metadata.csv".

• The entire data set can load into system memory
• The user has full permissions to access and edit the data
• There are no performance requirements to complete tasks quickly
• Any trouble making records should be dropped rather than manually fixed
• The environment has python >=3.6 and is Linux or Unix-Like

The file 'data/movie_metadata.csv' was downloaded from Kaggle on 2019-10-17 under the Open Database license.

Some records (eg with comma in title) may get dropped to avoid problems.

Given the constraints of time, not all features are implemented yet.

• Implement a small rest service. This would mostly just be a new type of views using the same controller but returning json responses instead.

• Testing. Unit testing for the business logic and integration/user acceptance testing for the views. The controller may need to be refactored a bit to be more friendly to unit testing though.

• More stat queries for actors, movies, and directors. Maybe a regression model or something fun.

• Load facebook likes for the actors/directors. Right now those fields are getting dropped.

• Replace the query in rank-personnel with an aggregation query. Right now the aggregation is done in code, which is very slow as it has to keep loading the movie records.