2022 / Day 29: Out of my comfort zone
I don’t feel comfortable around colors (as you might guess by the mostly b/w schemes I’ve been using this past challenge). So this is me trying to familiarize myself a bit with colors. And I still don’t feel comfortable. Please take me away from here.
But let’s get this over with. The following SQL draws three circles and colors
them based on the azimuth (from the central point) and the third one also
distance from the central point. All are divided into thee sectors.
The first one top left interpolates a single RGB color from min (0) to max (255)
in clockwise direction. The second one (top right) interpolates two colors in
mutually countering direction. So e.g. the downwards sector is a combination
of rgb(255,0,0) on the lefthand side and rgb(0,255,0) on the right with
rgb(128,128,0) somewhere in the middle. The other sectors are mixed
green and blue and blue and red.
The third one (middle bottom) combines all three colors - e.g. the downwards sector is calculated with red increasing clockwise, greens increasing counter-clockwise and blue increasing with distance from the center. And additionally a (360/255) degree turn with every blue DN increase.
with
center as (
select st_point(0,0) as geom, 1000 as w
),
data as (
select
s, width,
mod(s-1, 256) as q,
(s-1) / 256 as sector,
st_rotate(
st_polygonize(
st_addpoint(
st_addpoint(
st_linesubstring(
geom,
(s-1)::numeric/(3*256)::numeric,
s::numeric/(3*256)::numeric
),
center,
0
),
center,
-1
)
),
radians(-30)
) as geom
from (
select
st_boundary(
st_buffer(center.geom, center.w, 'quad_segs=80')
) as geom,
center.geom as center, center.w as width
from
center
) f, generate_series(1,3*256,1) s
group by s, width
order by s
),
verts as (
select
ring,
coalesce(
st_difference(
geom,
lag(geom) over(order by ring)
),
geom
) as geom
from (
select
s-1 as ring,
st_buffer(
center.geom,
(center.w::numeric/256.0) * s::numeric,
'quad_segs=80'
) as geom
from
center,
generate_series(1,256, 1) s
) a
),
onecolor as (
select
st_translate(
geom,
-1.0 * (width + 250.0),
(width + 250.0)
) as geom,
col[1] as red, col[2] as green, col[3] as blue, sector
from (
select
st_collectionextract(geom,3) as geom,
sector, width,
case
when sector=0 then
array[q,0, 0]
when sector=1 then
array[0,0,q]
when sector=2 then
array[0,q,0]
end as col
from
data
) f
),
twocolors as (
select
st_translate(
geom,
(width + 250.0),
(width + 250.0)
) as geom,
col[1] as red, col[2] as green, col[3] as blue, sector
from (
select
st_collectionextract(geom,3) as geom,
sector, width,
case
when sector=0 then
array[q,255-q, 0]
when sector=1 then
array[255-q,0,q]
when sector=2 then
array[0,q,255-q]
end as col
from
data
) f
),
threecolors as (
select
st_translate(
geom,
0.0,
-1.0 * (width/2.0 + 250.0)
) as geom,
col[1] as red, col[2] as green, col[3] as blue, sector
from (
select
st_rotate(
geom,
radians((360.0/255.0) * ring::numeric),
st_point(0,0)
) as geom,
sector, width,
case
when sector=0 then
array[q,255-q, ring]
when sector=1 then
array[255-q,ring,q]
when sector=2 then
array[ring,q,255-q]
end as col
from (
select
data.sector,
data.q,
data.width,
verts.ring,
st_intersection(x, verts.geom) as geom
from
verts,
data
join lateral
st_collectionextract(data.geom, 3) x on true
where
st_intersects(x, verts.geom)
) d
) f
)
select
row_number() over ()::int as oid,
geom, sector, red, green, blue
from (
select
geom, sector, red, green, blue
from
onecolor
union all
select
geom, sector, red, green, blue
from
twocolors
union all
select
geom, sector, red, green, blue
from
threecolors
) d
;