This is a simple D3 animation of flight shopping data for a single route, inspired by LiteBrite. A number of interesting patterns surface, like the downward pointing red triangles corresponding to expensive mid-week trips without a Saturday night stay, and small upward pointing triangles showing quick weekend trips. Travel gets cheaper after summer vacation (early September), but prices rise as departure approaches as shoppers tend to procrastinate and business travelers are late entrants to the market. Diamond-shaped blocks light up in response to flexible shopping queries like travel dates ± 3 days.
We use a hexagonal grid to show shopping results, with each cell representing specific travel dates. Departure date reads left to right for a sixty day period from Aug 1 to Oct 1. Each row shows length of stay with the top row indicating same day return (rare for Boston to London) and the bottom row being a three week stay. Columns diagonally down to the right are fixed departure dates, and columns diagonally up to the right are fixed return dates thanks to the identity departure date + length of stay = return date.
We animate flight shopping queries over time (displayed at top left) with time accelerated to about one shopping day per second. Bubble size indicates the volume of shopping (short advance and shorter stays tend to be more popular) and color shows price. Trips in the past are greyed out as shopping date exceeds departure date.