A one-liner with PDL's rvals:
pdl> p rvals(19,19,{Centre=>[5,4]}) <= 6
[
[0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
]
Edit: I know what you are thinking: what a dumb one-liner. What if frame-buffer is 1000x1000 px, but illuminated area just 10x10? Why calculate a million distances when clearly we can limit calculations to a small viewport (bounding box)? Ok, then:
pdl> $frame = zeroes 19,19
pdl> ($x,$y,$r,$xmax,$ymax) = (5,4,6,dims $frame)
pdl>
pdl> use List::Util
pdl> *min_ = \*List::Util::min
pdl> *max_ = \*List::Util::max
pdl>
pdl> ($llx,$lly,$urx,$ury,$cx,$cy) = (
> max_(0,$x-$r),
> max_(0,$y-$r),
> min_($xmax,$x+$r),
> min_($ymax,$y+$r),
> min_($r,$x),
> min_($r,$y))
pdl>
pdl> $viewport = $frame($llx:$urx, $lly:$ury)
pdl>
pdl> $viewport .= $viewport-> rvals({
> Center => [$cx,$cy],
> Squared => 1
> }) <= $r*$r
pdl>
pdl> p $frame
[
[0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0]
[0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
]
pdl>
Note two different kinds of assignment, for data-flow to work in PDL. This nice exercise reminded me when taking a square root multiple times was a no-no (was that so for /[2-4]87/? I think nowadays it's single CPU cycle anyway, no need to optimize (-?)), hence I used the rvals's option (but it's there for a reason, isn't it). I hope I didn't mess anything this time of night and it works for border cases, too.