What day is it today? code-challenge
Given a date y4y3y2y1-m2m1-d2d1, output the day in the week w. 0 as Sunday, 1 as Monday, and so on. You can assume the date is valid, and every year has 12 month, every Jan,Mar,May,Jul,Aug,Oct,Dec have 31 days, every Apr,Jun,Sep,Nov have 30 days, Feb have 29 ...
WAP in Java to input ten numbers and construct a max heap. After constructing max heap,
consider all the numbers that are at the leaf nodes of the heap. Input frequency of search of all
those numbers and create an optimal binary search tree using those numbers and their
frequency of search. Compu...
Task
Given a matrix of numbers \$M\$ with \$r\$ rows and \$c\$ columns, and the magnification factor \$n\$, build the matrix with \$rn\$ rows and \$cn\$ columns where the original elements are spaced \$n\$ units apart and the gaps are filled by piecewise linear interpolation:
$$
\begin{bmat...
x←?1e6⍴1000
a. ¯1+{≢⍵}⌸(⍳1000),x 1.00 (_1+#/.~(i.1000),x )
b. ¯1+{≢⍵}⌸(⍳1+⌈/x),x 1.62 (_1+#/.~(i.1+>./x),x)
that's the hint
@mypronounismonicareinstate Find the maximum of a vector of 1-byte ints without using multicore, vector instructions, loop unrolling, etc. Can you do it faster in C than the following?
perhaps a good idea would be to compute the OR of chars in blocks and if it's a subset of the current number to skip the entire block? (presumably, of 8 chars as an int64_t) (I guess I'll try that later) (I assume quitting once 255 was found is not fair, though) What unit of time is used in the benchmarks? (CPU cycles?)
I seem to be getting 175ms vs 202ms with `g++ -Os` (assuming -Os doesn't do cheating like unrolling; it probably shouldn't. but it seems to; this shouldn't happen), my code is `byte bmax(const unsigned char* u8x, int n) { const uint64_t* u64x = reinterpret_cast<const uint64_t*>(u8x); int i, j; byte max = 0; for(i = 0, j = 0; i < n; i += 8, j++) { /*"or" and "bitor" are keywords, of course */ uint64_t oR = u64x[j]; oR |= oR >> 32; oR |= oR >> 16; oR |= oR >> 8; if(byte(oR) > max) { for(int off = 0; off < 8; off++) max = std::max(max, u8x[i + off]); } } i -= 8; for(; i < n; i++) max = std::ma…
byte bmax(const unsigned char* u8x, int n)
{
const uint64_t* u64x = reinterpret_cast<const uint64_t*>(u8x);
int i, j;
byte max = 0;
for(i = 0, j = 0; i < n; i += 8, j++)
{
//"or" and "bitor" are keywords, of course
uint64_t oR = u64x[j];
oR |= oR >> 32;
oR |= oR >> 16;
oR |= oR >> 8;
if(byte(oR) > max)
{
for(int off = 0; off < 8; off++)
max = std::max(max, u8x[i + off]);
}
}
i -= 8;
for(; i < n; i++)
max = std::max(max, u8x[i]);
return max;
}
Bot Duels KOTH king-of-the-hill
Submissions
Submissions should be a JS function that takes the following parameters:
curr_x - the current x coordinate of your bot
curr_y - the current y coordinate of your bot
enemy_x - the current x coordinate of your opponent's bot
enemy_y - the current y co...
Yes; as I usually do in such cases, I xor all return values and printf the result. I was testing on 100000 test cases with 4096 random bytes. I think you missed a truncating cast in your code in the if(byte(oR) > max) line.
Lucky dice rolls
In pen and paper roleplaying games dice are used for various chance calculations. The usual way to describe a roll is \$n\textbf{d}k\$ where \$n\$ is the number of dice and \$k\$ is the number of faces on a die. For example \$3d6\$ means that you need to roll the classical 6-sid...
operations on larger numbers in terms of size has always been faster
than continuous operations on smaller numbers
I believe a lot comes into play. At first, I'd suspect partial register stalls, but I can't believe a compiler would make this mistake
maybe caching comes into play /shrug
and maybe the CPU pipeline is just optimized for computations on large numbers, this seems logical
that's a bit like how SIMD works, isn't it - larger vectors, the processor can do its job while the program isn't occupying other parts of the pipeline
anyway, why can't they use loop unrolling? (I can imagine some kind of special CPUs without SIMD, but can't imagine anything that can't use unrolled loops)
it's a typedef unsigned char byte;, a typedef I often (inconsistently) use because I'm coming from C# (over time, I started getting annoyed at its slowness)
@mypronounismonicareinstate to me it seems the requirements were there purely for the fun of the challenge (i.e. probably to prevent what yall're doing with oring :p)
@dzaima then the challenge looks like "golf me a fast fourier transform, but in Malbolge, to prove that ebe ebe"
for a long time I wanted to post a fastest-code Malbolge Unshackled interpreter hoping that someone will post an interpreter that will run my programs without taking exabytes of RAM and more processor cycles than the amount of particles observable universe squared :p
i'm more interested in what exactly was used for timing. if it's purely CPU cycles there are a couple things i'd try (depending on the platform, achieving 3 ops/item (including the jump to start) + O(1) wouldn't be hard, but it'd be slow otherwise), if it's time then the compiler using correct operations is pretty much all that matters
> Loop unrolling is a distasteful way to get speed-ups, and does not require much insight or skill. This view is of course depends on your point of view. Since I am posing the puzzle I get to make the rules. :-)
The context in which the puzzle is posed is a hint in itself. The "expert C programmers" to whom I posed the puzzle did not have the benefit of this hint.
@KrzysztofSzewczyk that's why i said the timing method and system configuration matters a lot (is constant memory bombarding a big issue, is the processor good at evaluating these types of instructions, etc), and otherwise it seems relatively related to the task at hand
@JohnDvorak Not only that, it also doesn't specify if the object is equidistant from the spots, and if if it is, whether it is directly above the great circle that connects the spots.
if I wanted to post a challenge regarding emulation of a certain automata, should I specify that a single storage cell should be unbounded? or leave it bounded and up to choice of the golfer?
You can calculate: the shortest path angle (planar equidistant case), longest path angle (planar case tangent at receiver), or the 3D case with both transmitter and receiver tangent
The challenge today is very simple. The task is to determine the TI calculator, for which the submitted datafile was made for.
The datafiles always start with a string **TI, the version (described below) and other data you can ignore.
Now, the versions you need to recognize are:
95* => TI-95
9...
the index-of challenge seems to be a nice project to do/
also you can still meet the requirements compiling with -O2, just add -fno-vectorize and -fno-unroll-loops, flags might be a bit different depending on compiler
iirc gcc doesnt enable vectorization until O3 but clang has it enabled at O1
Intro
Help! I'm stuck on a snow-covered mountain and I need to get down as fast as possible, preferably without dying. I have a map showing how high each part of the mountain is above the normal ground.
I am a novice skier, which means I like to stick to slopes that aren't too steep. I'm willin...
Task
Given a matrix of numbers \$M\$ with \$r\$ rows and \$c\$ columns, and the magnification factor \$n\$, build the matrix with \$rn\$ rows and \$cn\$ columns where the original elements are spaced \$n\$ units apart and the gaps are filled by piecewise linear interpolation:
$$
\begin{bmat...
Bot Duels KOTH king-of-the-hill
Submissions
Submissions should be a JS function that takes the following parameters:
curr_x - the current x coordinate of your bot
curr_y - the current y coordinate of your bot
enemy_x - the current x coordinate of your opponent's bot
enemy_y - the current y co...
The challenge today is very simple. The task is to determine the TI calculator, for which the submitted datafile was made for.
The datafiles always start with a string **TI, the version (described below) and other data you can ignore.
Now, the versions you need to recognize are:
95* => TI-95
9...
Intro
Help! I'm stuck on a snow-covered mountain and I need to get down as fast as possible, preferably without dying. I have a map showing how high each part of the mountain is above the normal ground.
I am a novice skier, which means I like to stick to slopes that aren't too steep. I'm willin...
