Even faster asin() was staring right at me (16bpp.net)
76 points by def-pri-pub 6 hours ago
srean 3 hours ago
A notable approximation of ~650 AD vintage, by Bhaskara is
ArcCos(x)= Π √((1-x)/(4+x)).
This tradition came to its own as Madhava school of mathematics from Kerala. https://en.wikipedia.org/wiki/Kerala_school_of_astronomy_and...
Note the approximation is for 0 < x < 1. For the range [-1, 0] Bhaskara used symmetry.
If I remember correctly, Aryabhatta had derived a rational approximation about a hundred years before this.
fhdkweig 5 hours ago
This is a followup of a different post from the same domain. 5 days ago, 134 comments https://news.ycombinator.com/item?id=47336111
coldcity_again 2 hours ago
I've been thinking about this since [1] the other day, but I still love how rotation by small angles lets you drop trig entirely.
Let α represent a roll rotation, and β a pitch rotation.
Let R(α) be:
( cos α sin α 0)
(-sin α cos α 0)
( 0 0 1)
(1 0 0 )
(0 cos β -sin β)
(0 sin β cos β)
R(β).R(α) = ( cos α sin α 0 )
((-sin α*cos β) (cos α*cos β) -sin β)
((-sin α*sin β) (cos α*sin β) cos β)
( 1 α 0)
(-α 1 -β)
( 0 β 1)
x' = x + αy
y' = y - αx - βz
z' = z + βy
coldcity_again an hour ago
If you just see the conclusion I think it's hard to immediately grok how rotation can arise from this.
This is a great technique for cheaply doing 3D starfields etc on 8-bit machines.
Look ma, no sine table!
srean an hour ago
A related interesting fact is that small angular motions compose almost like vectors, order does not matter (i.e. they are commutative). This makes differential kinematics easier to deal with when dealing with polar or cylindrical coordinate systems.
Large angular deflections while being linear transforms, do not in general commute.
It will spoil the linear relation in your elegant expression, but a slightly better approximation for cos for small θ is
1 - 0.5θ²ashdnazg 3 hours ago
No idea if it's not already optimised, but x2 could also be x*x and not just abs_x * abs_x, shifting the dependencies earlier.
jonasenordin 3 hours ago
I haven't kept up with C++ in a few years - what does constexpr do for local variables?
constexpr double a0 = 1.5707288;loeg 2 hours ago
It is required to be evaluated at compile time, and it's const.
An optimizing compiler might see through a non-constexpr declaration like 'double a0 = ...' or it might not. Constexpr is somewhat more explicit, especially with more complicated initializer expressions.
fancy_pantser 3 hours ago
The compiler can substitute the value how it sees fit. It's like #define, but type-safe and scoped.
Maybe it's folded into expressions, propagated through constant expressions, or used it in contexts that require compile-time constants (template parameters, array sizes, static_assert, other constexpr expressions).
I mean, not in this case of pi/2, where it's more about announcing semantics, but in general those are the purposes and uses.
gzread 2 hours ago
It can do this with const too or even a normal variable that just happens to not vary.
ErroneousBosh an hour ago
I'd like something like this in C or C++ quite honestly.
Something like a struct that I can say "this struct is global to the whole program and everyone can see it, but once this function exits those values are locked in". Maybe something like that one function is allowed to unlock and update it, but nowhere else.
Think in terms of storing a bunch of precomputed coefficients that are based on the samplerate of a system, where you really only need to set it up once on startup and it is unlikely to change during the application's running lifetime.
I feel like there probably is a way to do this, and if I was good at high level languages like C I'd know what it is. If you know, tell me what I'm not understanding ;-)
jaen 4 hours ago
Cool, although more ILP (instruction-level parallelism) might not necessarily be better on a modern GPU, which doesn't have much ILP, if any (instead it uses those resources to execute several threads in parallel).
That might explain why the original Cg (a GPU programming language) code did not use Estrin's, since at least the code in the post does add 1 extra op (squaring `abs_x`).
(AMD GPUs used to use VLIW (very long instruction word) which is "static" ILP).
jagged-chisel 3 hours ago
> It also gets in the way of elegance and truth.
That’s quite subjective. I happen to find trigonometry to be elegant and true.
I also agree that trigonometric functions lack efficiency in software.
def-pri-pub 3 hours ago
>> It also gets in the way of elegance and truth.
Where did that come from in the article?
jagged-chisel 2 hours ago
I revisited that article and ... now I have no idea. Maybe I stumbled into some other trig-related article and came back here. Or maybe this one had some A/B content going on?
The only thing I remember at this point is that I copied and pasted that sentence (I didn't type it.) Even search doesn't find the sentence anywhere but HN.
FEELmyAGI an hour ago
fatih-erikli-cg 5 hours ago
I think it is `atan` function. Sin is almost a lookup query.
xyzzyz 5 hours ago
On modern machines, looking things up can be slower than recomputing it, when the computation is simple. This is because the memory is much slower than the CPU, which means you can often compute something many times over before the answer from memory arrives.
somat 3 hours ago
Not just modern machines, the Nintendo64 was memory bound under most circumstances and as such many traditional optimizations (lookup tables, unrolling loops) can be slower on the N64. The unrolling loops case is interesting. Because the cpu has to fetch more instructions this puts more strain on the memory bus.
If curious, On a N64 the graphics chip is also the memory controller so every thing the cpu can do to stay off the memory bus has an additive effect allowing the graphics to do more graphics. This is also why the n64 has weird 9-bit ram, it is so they could use a 18-bit pixel format, only taking two bytes per pixel, for cpu requests the memory controller ignored the 9th bit, presenting a normal 8 bit byte.
They were hoping that by having high speed memory, 250 mHz, the cpu ran at 90mHz, it could provide for everyone and it did ok, there are some very impressive games on the n64. but on most of them the cpu is running fairly light, gotta stay off that memory bus.
https://www.youtube.com/watch?v=xFKFoGiGlXQ (Kaze Emanuar: Finding the BEST sine function for Nintendo 64)
gzread 2 hours ago
ErroneousBosh 37 minutes ago
connicpu 4 hours ago
Unless your lookup table is small enough to only use a portion of your L1 cache and you're calling it so much that the lookup table is never evicted :)
fatih-erikli-cg 5 hours ago
It may be, especially when it comes to unnecessary cache. But I think `atan` is almost a brute force. Lookup is nothing comparing to that.
Sin/cos must be borders of sqrt(x²+y²). It is also cached indeed.
groundzeros2015 4 hours ago
BigTTYGothGF an hour ago
thomasahle 5 hours ago
Did you try polynomial preprocessing methods, like Knuth's and Estrin's methods? https://en.wikipedia.org/wiki/Polynomial_evaluation#Evaluati... they let you compute polynomials with half the multiplications of Horner's method, and I used them in the past to improve the speed of the exponential function in Boost.
kzrdude 4 hours ago
yes, Estrin's method is the update
thomasahle 4 hours ago
Sorry, I said that wrong. Estrin's doesn't reduce the number of multiplications.