Comment by amiga386

Boeing vs Airbus:

    count() AS total,
    sum(desc LIKE 'BOEING%') AS boeing,
    sum(desc LIKE 'AIRBUS%') AS airbus,
    sum(NOT (desc LIKE 'BOEING%' OR desc LIKE 'AIRBUS%')) AS other,

    greatest(1000000 DIV {sampling:UInt32} DIV zoom_factor, total) AS max_total,
    greatest(1000000 DIV {sampling:UInt32} DIV zoom_factor, boeing) AS max_boeing,
    greatest(1000000 DIV {sampling:UInt32} DIV zoom_factor, airbus) AS max_airbus,
    greatest(1000000 DIV {sampling:UInt32} DIV zoom_factor, other) AS max_other,

    pow(total / max_total, 1/5) AS transparency,

    255 * (1 + transparency) / 2 AS alpha,
    pow(boeing, 1/5) * 256 DIV (1 + pow(max_boeing, 1/5)) AS red,
    pow(airbus, 1/5) * 256 DIV (1 + pow(max_airbus, 1/5)) AS green,
    pow(other, 1/5) * 256 DIV (1 + pow(max_other, 1/5)) AS blue

    SELECT round(red)::UInt8, round(green)::UInt8, round(blue)::UInt8, round(alpha)::UInt8

The redder the pixel, the more Boeing planes there.

The greener the pixel, the more Airbus planes there.

The bluer the pixel, the more non-Boeing/Airbus planes there.

The less transparent the pixel, the more planes in total.

White means all planes fly there, yellow means Boeing and Airbus dominate, red means Boeing dominates, green means Airbus dominates, cyan means Airbus+others, magenta means Boeing+others, etc.