Using scheme I have the need to use the following function. (All args are natural numbers [0, inf) )
(define safe-div
(lambda (num denom safe)
(if (zero? denom)
safe
(div num denom))))
However, this function is called quite often and is not performing well enough (speed wise).
Is there a more efficient way of implementing the desired behavior (integer division of num and denom, return safe value if denom is zero)?
Notes, I am using Chez Scheme, however this is being used in a library that imports rnrs only, not full Chez.
For maximum performance, you need to get as close to the silicon as possible. Adding safety checks like this isn't going to do it, unless they get just-in-time compiled into ultra-efficient machine code by the scheme system.
I see two options. One is to create a native (i.e. foreign) implementation in C (or assembly) and invoke that. That might not be compatible with packaging it as a lambda, but then again, the dynamic nature of lambdas leads to notational efficiency but not necessarily runtime efficiency. (Function pointers excepted, there's a reason lambda expressions are not present in C, despite being many years older.) If you go this route, it might be best to take a step back and see if the larger processing of which safe-div is a part should be taken native. There's little point in speeding up the division at the center of a loop if everything around it is still slow.
Assuming that division by zero is expected to be rare, another approach is to just use div and hope its implementation is fast. Yes, this can lead to division by zero, but when it comes to speed, sometimes it is better to beg forgiveness than to ask permission. In other words, skip the checking before the division and just do it. If it fails, the scheme runtime should catch the division by zero fault and you can install an exception handler for it. This leads to slower code in the exceptional case and faster code in the normal case. Hopefully this tradeoff works out to a performance win.
Lastly, depending on what you are dividing by, it might be faster to multiply by the reciprocal than to perform an actual division. This requires fast reciprocal computation or revising earlier computations to yield a reciprocal directly. Since you are dealing with integers, the reciprocal would be stored in fixed-point, which is essentially 2^32 * 1/denom. Multiply this by num and shift right by 32 bits to get the quotient. This works out to a win because more processors these days have single cycle multiply instructions, but division executes in a loop on the chip, which is much slower. This might be overkill for your needs, but could be useful at some point.
Related
//My question was so long So I reduced.
In scheme, user-made procedures consume more time than built-in procedures?
(If both's functions are same)
//This is my short version question.
//Below is long long version question.
EX 1.23 is problem(below), why the (next) procedure isn't twice faster than (+ 1)?
This is my guess.
reason 1 : (next) contains 'if' (special-form) and it consumes time.
reason 2 : function call consumes time.
http://community.schemewiki.org/?sicp-ex-1.23 says reason 1 is right.
And I want to know reason 2 is also right.
SO I rewrote the (next) procedure. I didn't use 'if' and checked the number divided by 2 just once before use (next)(so (next) procedure only do + 2). And I remeasured the time. It was more fast than before BUT still not 2. SO I rewrote again. I changed (next) to (+ num 2). Finally It became 2 or almost 2. And I thought why. This is why I guess the 'reason 2'. I want to know what is correct answer.
ps. I'm also curious about why some primes are being tested (very?) fast than others? It doesn't make sense because if a number n is prime, process should see from 2 to sqrt(n). But some numbers are tested faster. Do you know why some primes are tested faster?
Exercise 1.23. The smallest-divisor procedure shown at the start of this section does lots of needless testing: After it checks to see if the number is divisible by 2 there is no point in checking to see if it is divisible by any larger even numbers. This suggests that the values used for test-divisor should not be 2, 3, 4, 5, 6, ..., but rather 2, 3, 5, 7, 9, .... To implement this change, define a procedure next that returns 3 if its input is equal to 2 and otherwise returns its input plus 2. Modify the smallest-divisor procedure to use (next test-divisor) instead of (+ test-divisor 1). With timed-prime-test incorporating this modified version of smallest-divisor, run the test for each of the 12 primes found in exercise 1.22. Since this modification halves the number of test steps, you should expect it to run about twice as fast. Is this expectation confirmed? If not, what is the observed ratio of the speeds of the two algorithms, and how do you explain the fact that it is different from 2?
Where the book is :
http://mitpress.mit.edu/sicp/full-text/book/book-Z-H-11.html#%_sec_1.2.6
Your long ans short questions are actually addressing two different problems.
EX 1.23 is problem(below), why the (next) procedure isn't twice faster than (+ 1)?
You provide two possible reasons to explain the relative lack of speed up (and 30% speed gain is already a good achievement):
The apparent use of a function next instead of a simple arithmetic expression (as I understand it from your explanations),
the inclusion of a test in that very next function,
The first reason is an illusion: (+ 1) is a function incrementing its argument, so in both cases there is a function call (although the increment function is certainly a builtin one, a point which is addressed by your other question).
The second reason is indeed relevant: any test in a block of code will introduce a potential redirection in the code flow (a jump from the current executing instruction to some other address in the program), which may incur a delay. Note that this is analogous to a function call, which will also induce an address jump, so both reasons actually resolve to only one potential cause.
Regarding your short question, builtin functions are indeed usually faster, because the compiler is able to apply a special treatment to them in certain cases. This is due to the facts that:
knowing the semantics of the builtins, compiler designers are able to include special rules pertaining to the algebraic properties of these builtins, and for instance fuse successive incréments in a single call, or suppress a combination of increment and decrement called in sequence.
A builtin function call, when not optimised away, will be converted into a native machine code function call, which may not have to abide to all the calling convention rules. If your scheme compiler produce machine code from the source, then there might be only a marginal gain, but if it produce so called bytecode, the gain might be quite substantial, since user written functions will be translated to that bytecode format, and still require some form of interpretation. If you are only using an interpreter, then the gain is even more important.
I believe this is highly implementation and setting dependent. In many implementations there are different kinds of optimizations and in some there are none. To get the best performance you may need to compile your code or use settings to reduce debug information / stack traces. Getting the best performance in one implementation can worsen the performance in another.
Primitive procedures are usually compiled to be native and in some implementations, like ikarus, it's even inlined. (When you do (map car lst) ikarus changes it to (map (lambda (x) (car x)) lst) since car isn't a procedure) Lambdas are supposed to be cheap.. Remember many scheme implementations change your code to CPS and that is one procedure call for each expression in the body of a procedure call. It will never be as fast as machine code since it needs to do load closure variables.
To check which of the two options which are correct for your implementation make next do the same as it originally did. eg. no if but just increment the argument. The difference now is the extra call and nothing else. Then you can inline next by writing it's code directly in your procedure and substituting arguments for the operands. Is it still slower, then it's if. you need to run the tests several times, preferably with large enough number of primes to produce that it runs for a minute or so. Use time or similar in both the Scheme implementations to get the differences in ms. I use unix time command as well to see how the OS reflects on it.
You should also test to see if you get the same reason in some other implementation. It's not like it's not enough Scheme implementations out there so know yourself out! The differences between them might amaze you. I always use racket (raco exe source to make executable) and Ikarus. WHen doing a large test I include Chicken, Gambit and Chibi.
Recently, i heard that % operator is costly in terms of time.
So, the question is that, is there a way to find the remainder faster?
Also your help will be appreciated if anyone can tell the difference in the execution of % / * + - operations.
In some cases where you're using power-of-2 divisors you can do better with roll-your-own techniques for calculating remainder, but generally a halfway decent compiler will do the best job possible with variable divisors, or "odd" divisors that don't fit any pattern.
Note that a few CPUs don't even have a multiply operation, and so (on those) multiply is quite slow vs add (at least 64x for a 32-bit multiply). (But a smart compiler may improve on this if the multiplier is a literal.) A slightly larger number do not have a divide operation or have a pretty slow one. (On a CPU with a fast multiplier multiply may only be on the order of 4 times slower than add, but on "normal" hardware it's 16-32 times slower for a 32 bit operation. Divide is inherently 2-4x slower than multiply, but can be much slower on some hardware.)
The remainder operation is rarely implemented in hardware, and normally A % B maps to something along the lines of A - ((A / B) * B) (a few extra operations may be required to assure the proper sign, et al).
(I learned about this stuff while microprogramming the instruction set for the SUMC computer for RCA/NASA back in the early 70s.)
No, the compiler is going to implement % in the most efficient way possible.
In terms of speed, + and - are the fastest (and are equally fast, generally done by the same hardware).
*, /, and % are much slower. Multiplication is basically done by the method you learn in grade school- multiply the first number by every digit in the second number and add the results. With some hacks made possible by binary. As of a few years ago, multiply was 3x slower than add. Division should be similar to multiply. Remainder is similar to division (in fact it generally calculates both at once).
Exact differences depend on the CPU type and exact model. You'd need to look up the latencies in the CPU spec sheets for your particular machine.
Recently I realized I have been doing too much branching without caring the negative impact on performance it had, therefore I have made up my mind to attempt to learn all about not branching. And here is a more extreme case, in attempt to make the code to have as little branch as possible.
Hence for the code
if(expression)
A = C; //A and C have to be the same type here obviously
expression can be A == B, or Q<=B, it could be anything that resolve to true or false, or i would like to think of it in term of the result being 1 or 0 here
I have come up with this non branching version
A += (expression)*(C-A); //Edited with thanks
So my question would be, is this a good solution that maximize efficiency?
If yes why and if not why?
Depends on the compiler, instruction set, optimizer, etc. When you use a boolean expression as an int value, e.g., (A == B) * C, the compiler has to do the compare, and the set some register to 0 or 1 based on the result. Some instruction sets might not have any way to do that other than branching. Generally speaking, it's better to write simple, straightforward code and let the optimizer figure it out, or find a different algorithm that branches less.
Jeez, no, don't do that!
Anyone who "penalize[s] [you] a lot for branching" would hopefully send you packing for using something that awful.
How is it awful, let me count the ways:
There's no guarantee you can multiply a quantity (e.g., C) by a boolean value (e.g., (A==B) yields true or false). Some languages will, some won't.
Anyone casually reading it is going observe a calculation, not an assignment statement.
You're replacing a comparison, and a conditional branch with two comparisons, two multiplications, a subtraction, and an addition. Seriously non-optimal.
It only works for integral numeric quantities. Try this with a wide variety of floating point numbers, or with an object, and if you're really lucky it will be rejected by the compiler/interpreter/whatever.
You should only ever consider doing this if you had analyzed the runtime properties of the program and determined that there is a frequent branch misprediction here, and that this is causing an actual performance problem. It makes the code much less clear, and its not obvious that it would be any faster in general (this is something you would also have to measure, under the circumstances you are interested in).
After doing research, I came to the conclusion that when there are bottleneck, it would be good to include timed profiler, as these kind of codes are usually not portable and are mainly used for optimization.
An exact example I had after reading the following question below
Why is it faster to process a sorted array than an unsorted array?
I tested my code on C++ using that, that my implementation was actually slower due to the extra arithmetics.
HOWEVER!
For this case below
if(expression) //branched version
A += C;
//OR
A += (expression)*(C); //non-branching version
The timing was as of such.
Branched Sorted list was approximately 2seconds.
Branched unsorted list was aproximately 10 seconds.
My implementation (whether sorted or unsorted) are both 3seconds.
This goes to show that in an unsorted area of bottleneck, when we have a trivial branching that can be simply replaced by a single multiplication.
It is probably more worthwhile to consider the implementation that I have suggested.
** Once again it is mainly for the areas that is deemed as the bottleneck **
I'm programming a PLC with some legacy software (RSLogix 500, don't ask) and it does not natively support a modulus operation, but I need one. I do not have access to: modulus, integer division, local variables, a truncate operation (though I can hack it with rounding). Furthermore, all variables available to me are laid out in tables sorted by data type. Finally, it should work for floating point decimals, for example 12345.678 MOD 10000 = 2345.678.
If we make our equation:
dividend / divisor = integer quotient, remainder
There are two obvious implementations.
Implementation 1:
Perform floating point division: dividend / divisor = decimal quotient. Then hack together a truncation operation so you find the integer quotient. Multiply it by the divisor and find the difference between the dividend and that, which results in the remainder.
I don't like this because it involves a bunch of variables of different types. I can't 'pass' variables to a subroutine, so I just have to allocate some of the global variables located in multiple different variable tables, and it's difficult to follow. Unfortunately, 'difficult to follow' counts, because it needs to be simple enough for a maintenance worker to mess with.
Implementation 2:
Create a loop such that while dividend > divisor divisor = dividend - divisor. This is very clean, but it violates one of the big rules of PLC programming, which is to never use loops, since if someone inadvertently modifies an index counter you could get stuck in an infinite loop and machinery would go crazy or irrecoverably fault. Plus loops are hard for maintenance to troubleshoot. Plus, I don't even have looping instructions, I have to use labels and jumps. Eww.
So I'm wondering if anyone has any clever math hacks or smarter implementations of modulus than either of these. I have access to + - * /, exponents, sqrt, trig functions, log, abs value, and AND/OR/NOT/XOR.
How many bits are you dealing with? You could do something like:
if dividend > 32 * divisor dividend -= 32 * divisor
if dividend > 16 * divisor dividend -= 16 * divisor
if dividend > 8 * divisor dividend -= 8 * divisor
if dividend > 4 * divisor dividend -= 4 * divisor
if dividend > 2 * divisor dividend -= 2 * divisor
if dividend > 1 * divisor dividend -= 1 * divisor
quotient = dividend
Just unroll as many times as there are bits in dividend. Make sure to be careful about those multiplies overflowing. This is just like your #2 except it takes log(n) instead of n iterations, so it is feasible to unroll completely.
If you don't mind overly complicating things and wasting computer time you can calculate modulus with periodic trig functions:
atan(tan(( 12345.678 -5000)*pi/10000))*10000/pi+5000 = 2345.678
Seriously though, subtracting 10000 once or twice (your "implementation 2") is better. The usual algorithms for general floating point modulus require a number of bit-level manipulations that are probably unfeasible for you. See for example http://www.netlib.org/fdlibm/e_fmod.c (The algorithm is simple but the code is complex because of special cases and because it is written for IEEE 754 double precision numbers assuming there is no 64-bit integer type)
This all seems completely overcomplicated. You have an encoder index that rolls over at 10000 and objects rolling along the line whose positions you are tracking at any given point. If you need to forward project stop points or action points along the line, just add however many inches you need and immediately subtract 10000 if your target result is greater than 10000.
Alternatively, or in addition, you always get a new encoder value every PLC scan. In the case where the difference between the current value and last value is negative you can energize a working contact to flag the wrap event and make appropriate corrections for any calculations on that scan. (**or increment a secondary counter as below)
Without knowing more about the actual problem it is hard to suggest a more specific solution but there are certainly better solutions. I don't see a need for MOD here at all. Furthermore, the guys on the floor will thank you for not filling up the machine with obfuscated wizard stuff.
I quote :
Finally, it has to work for floating point decimals, for example
12345.678 MOD 10000 = 2345.678
There is a brilliant function that exists to do this - it's a subtraction. Why does it need to be more complicated than that? If your conveyor line is actually longer than 833 feet then roll a second counter that increments on a primary index roll-over until you've got enough distance to cover the ground you need.
For example, if you need 100000 inches of conveyor memory you can have a secondary counter that rolls over at 10. Primary encoder rollovers can be easily detected as above and you increment the secondary counter each time. Your working encoder position, then, is 10000 times the counter value plus the current encoder value. Work in the extended units only and make the secondary counter roll over at whatever value you require to not lose any parts. The problem, again, then reduces to a simple subtraction (as above).
I use this technique with a planetary geared rotational part holder, for example. I have an encoder that rolls over once per primary rotation while the planetary geared satellite parts (which themselves rotate around a stator gear) require 43 primary rotations to return to an identical starting orientation. With a simple counter that increments (or decrements, depending on direction) at the primary encoder rollover point it gives you a fully absolute measure of where the parts are at. In this case, the secondary counter rolls over at 43.
This would work identically for a linear conveyor with the only difference being that a linear conveyor can go on for an infinite distance. The problem then only needs to be limited by the longest linear path taken by the worst-case part on the line.
With the caveat that I've never used RSLogix, here is the general idea (I've used generic symbols here and my syntax is probably a bit wrong but you should get the idea)
With the above, you end up with a value ENC_EXT which has essentially transformed your encoder from a 10k inch one to a 100k inch one. I don't know if your conveyor can run in reverse, if it can you would need to handle the down count also. If the entire rest of your program only works with the ENC_EXT value then you don't even have to worry about the fact that your encoder only goes to 10k. It now goes to 100k (or whatever you want) and the wraparound can be handled with a subtraction instead of a modulus.
Afterword :
PLCs are first and foremost state machines. The best solutions for PLC programs are usually those that are in harmony with this idea. If your hardware is not sufficient to fully represent the state of the machine then the PLC program should do its best to fill in the gaps for that missing state information with the information it has. The above solution does this - it takes the insufficient 10000 inches of state information and extends it to suit the requirements of the process.
The benefit of this approach is that you now have preserved absolute state information, not just for the conveyor, but also for any parts on the line. You can track them forward and backward for troubleshooting and debugging and you have a much simpler and clearer coordinate system to work with for future extensions. With a modulus calculation you are throwing away state information and trying to solve individual problems in a functional way - this is often not the best way to work with PLCs. You kind of have to forget what you know from other programming languages and work in a different way. PLCs are a different beast and they work best when treated as such.
You can use a subroutine to do exactly what you are talking about. You can tuck the tricky code away so the maintenance techs will never encounter it. It's almost certainly the easiest for you and your maintenance crew to understand.
It's been a while since I used RSLogix500, so I might get a couple of terms wrong, but you'll get the point.
Define a Data File each for your floating points and integers, and give them symbols something along the lines of MOD_F and MOD_N. If you make these intimidating enough, maintenance techs leave them alone, and all you need them for is passing parameters and workspace during your math.
If you really worried about them messing up the data tables, there are ways to protect them, but I have forgotten what they are on a SLC/500.
Next, defined a subroutine, far away numerically from the ones in use now, if possible. Name it something like MODULUS. Again, maintenance guys almost always stay out of SBRs if they sound like programming names.
In the rungs immediately before your JSR instruction, load the variables you want to process into the MOD_N and MOD_F Data Files. Comment these rungs with instructions that they load data for MODULUS SBR. Make the comments clear to anyone with a programming background.
Call your JSR conditionally, only when you need to. Maintenance techs do not bother troubleshooting non-executing logic, so if your JSR is not active, they will rarely look at it.
Now you have your own little walled garden where you can write your loop without maintenance getting involved with it. Only use those Data Files, and don't assume the state of anything but those files is what you expect. In other words, you cannot trust indirect addressing. Indexed addressing is OK, as long as you define the index within your MODULUS JSR. Do not trust any incoming index. It's pretty easy to write a FOR loop with one word from your MOD_N file, a jump and a label. Your whole Implementation #2 should be less than ten rungs or so. I would consider using an expression instruction or something...the one that lets you just type in an expression. Might need a 504 or 505 for that instruction. Works well for combined float/integer math. Check the results though to make sure the rounding doesn't kill you.
After you are done, validate your code, perfectly if possible. If this code ever causes a math overflow and faults the processor, you will never hear the end of it. Run it on a simulator if you have one, with weird values (in case they somehow mess up the loading of the function inputs), and make sure the PLC does not fault.
If you do all that, no one will ever even realize you used regular programming techniques in the PLC, and you will be fine. AS LONG AS IT WORKS.
This is a loop based on the answer by #Keith Randall, but it also maintains the result of the division by substraction. I kept the printf's for clarity.
#include <stdio.h>
#include <limits.h>
#define NBIT (CHAR_BIT * sizeof (unsigned int))
unsigned modulo(unsigned dividend, unsigned divisor)
{
unsigned quotient, bit;
printf("%u / %u:", dividend, divisor);
for (bit = NBIT, quotient=0; bit-- && dividend >= divisor; ) {
if (dividend < (1ul << bit) * divisor) continue;
dividend -= (1ul << bit) * divisor;
quotient += (1ul << bit);
}
printf("%u, %u\n", quotient, dividend);
return dividend; // the remainder *is* the modulo
}
int main(void)
{
modulo( 13,5);
modulo( 33,11);
return 0;
}
20-30 years ago arithmetic operations like division were one of the most costly operations for CPUs. Saving one division in a piece of repeatedly called code was a significant performance gain. But today CPUs have fast arithmetic operations and since they heavily use instruction pipelining, conditionals can disrupt efficient execution. If I want to optimize code for speed, should I prefer arithmetic operations in favor of conditionals?
Example 1
Suppose we want to implement operations modulo n. What will perform better:
int c = a + b;
result = (c >= n) ? (c - n) : c;
or
result = (a + b) % n;
?
Example 2
Let's say we're converting 24-bit signed numbers to 32-bit. What will perform better:
int32_t x = ...;
result = (x & 0x800000) ? (x | 0xff000000) : x;
or
result = (x << 8) >> 8;
?
All the low hanging fruits are already picked and pickled by authors of compilers and guys who build hardware. If you are the kind of person who needs to ask such question, you are unlikely to be able to optimize anything by hand.
While 20 years ago it was possible for a relatively competent programmer to make some optimizations by dropping down to assembly, nowadays it is the domain of experts, specializing in the target architecture; also, optimization requires not only knowing the program, but knowing the data it will process. Everything comes down to heuristics, tests under different conditions etc.
Simple performance questions no longer have simple answers.
If you want to optimise for speed, you should just tell your compiler to optimise for speed. Modern compilers will generally outperform you in this area.
I've sometimes been surprised trying to relate assembly code back to the original source for this very reason.
Optimise your source code for readability and let the compiler do what it's best at.
I expect that in example #1, the first will perform better. The compiler will probably apply some bit-twiddling trick to avoid a branch. But you're taking advantage of knowledge that it's extremely unlikely that the compiler can deduce: namely that the sum is always in the range [0:2*n-2] so a single subtraction will suffice.
For example #2, the second way is both faster on modern CPUs and simpler to follow. A judicious comment would be appropriate in either version. (I wouldn't be surprised to see the compiler convert the first version into the second.)