My topic is about movie genres, I want to explore the people’s preference to rent dvds on special day or season in Seattle. People might rent dvds and see movies to commemorate something, especially on special days, such as New Year, Valentine's Day, Easter, Christmas, and so on. I want to explore if people would love to see different kind movies on special day.
I filtered out popular films(top 10 ) for each type of films from the dataset of most popular films in Seattle library and figure out the pattern of days that typical movies of different type were out of library.
title, date, COUNT(date) AS Counts, MIN(duration) AS days
title, DATE(cout) AS date,
TIMESTAMPDIFF(HOUR, cout, cin) / 24.0 AS duration
itemtype like 'acdvd' AND
( title like "Into the wild" OR
title like "Burn after reading" OR
title like "Michael Clayton" OR
title like "No country for old men" OR
title like "reader" OR
title like "Atonement" OR
title like "Doubt" OR
title like "curious case of Benjamin Button" OR
title like "Slumdog millionaire" OR
title like "Vicky Cristina Barcelona" OR…
GROUP BY date, title, duration
) AS charts
GROUP BY title, date, days
ORDER BY Counts DESC
After data acquisition, I sorted my movie dataset into 6 genres according to IMDB reference.
In my project, I use the shape of CD as outlook of the data visualization. All of the movies will be represented on the surface of the “CDs” using the shape of box with different colors(red represent action…). I use the edge of cds to measure dates. There are 12 months in total. I have tried the linear measure to see the result but the outlook of circle shape could present all on the screen so it will be most effective way to compare data everyday. I use controller to filter out data on special day of season, so the viewer can see clear comparison of genres on special period. The Radius distance present the counts of renting of each movies. The counts increase from the centre of a circle to the edge of a circle. Meanwhile I use animation of jumping speed and length of box to measure the renting duration of each movies. The shorter of length and fast speed of jump, the shorter time of each dvds being out of library at least. In this way, I could explore which movie genre will be kept for long time to memorize or repeating, which is more like Popcorn movie. There will be 10 years of renting records and I use vertical dimension to present for each year. And I use the camera to travel through the year for the clear look of data.
Design Representation and Screenshots
Here are the results of visualization. When we move through year, we could see that the circle getting smaller, because the renting counts is decrease through year. And the Radioactive object getting more continuous because the rental duration getting longer
I draw a line to connect famous film(rental time > 5/day) in case of the length of box may block the movies with more counting times.I found people preferred to rent Romance and Drama film from April to August which is around spring and summer; Action films from autumn to winter;Comedies all over the year;Fantasy films seem have no clear regulations.
All work is developed within Processing: