I understand what the unsigned right shift operator ">>>" in Java does, but why do we need it, and why do we not need a corresponding unsigned left shift operator?
The >>> operator lets you treat int and long as 32- and 64-bit unsigned integral types, which are missing from the Java language.
This is useful when you shift something that does not represent a numeric value. For example, you could represent a black and white bit map image using 32-bit ints, where each int encodes 32 pixels on the screen. If you need to scroll the image to the right, you would prefer the bits on the left of an int to become zeros, so that you could easily put the bits from the adjacent ints:
int shiftBy = 3;
int[] imageRow = ...
int shiftCarry = 0;
// The last shiftBy bits are set to 1, the remaining ones are zero
int mask = (1 << shiftBy)-1;
for (int i = 0 ; i != imageRow.length ; i++) {
// Cut out the shiftBits bits on the right
int nextCarry = imageRow & mask;
// Do the shift, and move in the carry into the freed upper bits
imageRow[i] = (imageRow[i] >>> shiftBy) | (carry << (32-shiftBy));
// Prepare the carry for the next iteration of the loop
carry = nextCarry;
}
The code above does not pay attention to the content of the upper three bits, because >>> operator makes them
There is no corresponding << operator because left-shift operations on signed and unsigned data types are identical.
>>> is also the safe and efficient way of finding the rounded mean of two (large) integers:
int mid = (low + high) >>> 1;
If integers high and low are close to the the largest machine integer, the above will be correct but
int mid = (low + high) / 2;
can get a wrong result because of overflow.
Here's an example use, fixing a bug in a naive binary search.
Basically this has to do with sign (numberic shifts) or unsigned shifts (normally pixel related stuff).
Since the left shift, doesn't deal with the sign bit anyhow, it's the same thing (<<< and <<)...
Either way I have yet to meet anyone that needed to use the >>>, but I'm sure they are out there doing amazing things.
As you have just seen, the >> operator automatically fills the
high-order bit with its previous contents each time a shift occurs.
This preserves the sign of the value. However, sometimes this is
undesirable. For example, if you are shifting something that does not
represent a numeric value, you may not want sign extension to take
place. This situation is common when you are working with pixel-based
values and graphics. In these cases you will generally want to shift a
zero into the high-order bit no matter what its initial value was.
This is known as an unsigned shift. To accomplish this, you will use
java’s unsigned, shift-right operator,>>>, which always shifts zeros
into the high-order bit.
Further reading:
http://henkelmann.eu/2011/02/01/java_the_unsigned_right_shift_operator
http://www.java-samples.com/showtutorial.php?tutorialid=60
The signed right-shift operator is useful if one has an int that represents a number and one wishes to divide it by a power of two, rounding toward negative infinity. This can be nice when doing things like scaling coordinates for display; not only is it faster than division, but coordinates which differ by the scale factor before scaling will differ by one pixel afterward. If instead of using shifting one uses division, that won't work. When scaling by a factor of two, for example, -1 and +1 differ by two, and should thus differ by one afterward, but -1/2=0 and 1/2=0. If instead one uses signed right-shift, things work out nicely: -1>>1=-1 and 1>>1=0, properly yielding values one pixel apart.
The unsigned operator is useful either in cases where either the input is expected to have exactly one bit set and one will want the result to do so as well, or in cases where one will be using a loop to output all the bits in a word and wants it to terminate cleanly. For example:
void processBitsLsbFirst(int n, BitProcessor whatever)
{
while(n != 0)
{
whatever.processBit(n & 1);
n >>>= 1;
}
}
If the code were to use a signed right-shift operation and were passed a negative value, it would output 1's indefinitely. With the unsigned-right-shift operator, however, the most significant bit ends up being interpreted just like any other.
The unsigned right-shift operator may also be useful when a computation would, arithmetically, yield a positive number between 0 and 4,294,967,295 and one wishes to divide that number by a power of two. For example, when computing the sum of two int values which are known to be positive, one may use (n1+n2)>>>1 without having to promote the operands to long. Also, if one wishes to divide a positive int value by something like pi without using floating-point math, one may compute ((value*5468522205L) >>> 34) [(1L<<34)/pi is 5468522204.61, which rounded up yields 5468522205]. For dividends over 1686629712, the computation of value*5468522205L would yield a "negative" value, but since the arithmetically-correct value is known to be positive, using the unsigned right-shift would allow the correct positive number to be used.
A normal right shift >> of a negative number will keep it negative. I.e. the sign bit will be retained.
An unsigned right shift >>> will shift the sign bit too, replacing it with a zero bit.
There is no need to have the equivalent left shift because there is only one sign bit and it is the leftmost bit so it only interferes when shifting right.
Essentially, the difference is that one preserves the sign bit, the other shifts in zeros to replace the sign bit.
For positive numbers they act identically.
For an example of using both >> and >>> see BigInteger shiftRight.
In the Java domain most typical applications the way to avoid overflows is to use casting or Big Integer, such as int to long in the previous examples.
int hiint = 2147483647;
System.out.println("mean hiint+hiint/2 = " + ( (((long)hiint+(long)hiint)))/2);
System.out.println("mean hiint*2/2 = " + ( (((long)hiint*(long)2)))/2);
BigInteger bhiint = BigInteger.valueOf(2147483647);
System.out.println("mean bhiint+bhiint/2 = " + (bhiint.add(bhiint).divide(BigInteger.valueOf(2))));
Related
What is the difference between objective-c C primitive numbers? I know what they are and how to use them (somewhat), but I'm not sure what the capabilities and uses of each one is. Could anyone clear up which ones are best for some scenarios and not others?
int
float
double
long
short
What can I store with each one? I know that some can store more precise numbers and some can only store whole numbers. Say for example I wanted to store a latitude (possibly retrieved from a CLLocation object), which one should I use to avoid loosing any data?
I also noticed that there are unsigned variants of each one. What does that mean and how is it different from a primitive number that is not unsigned?
Apple has some interesting documentation on this, however it doesn't fully satisfy my question.
Well, first off types like int, float, double, long, and short are C primitives, not Objective-C. As you may be aware, Objective-C is sort of a superset of C. The Objective-C NSNumber is a wrapper class for all of these types.
So I'll answer your question with respect to these C primitives, and how Objective-C interprets them. Basically, each numeric type can be placed in one of two categories: Integer Types and Floating-Point Types.
Integer Types
short
int
long
long long
These can only store, well, integers (whole numbers), and are characterized by two traits: size and signedness.
Size means how much physical memory in the computer a type requires for storage, that is, how many bytes. Technically, the exact memory allocated for each type is implementation-dependendant, but there are a few guarantees: (1) char will always be 1 byte (2) sizeof(short) <= sizeof(int) <= sizeof(long) <= sizeof(long long).
Signedness means, simply whether or not the type can represent negative values. So a signed integer, or int, can represent a certain range of negative or positive numbers (traditionally –2,147,483,648 to 2,147,483,647), and an unsigned integer, or unsigned int can represent the same range of numbers, but all positive (0 to 4,294,967,295).
Floating-Point Types
float
double
long double
These are used to store decimal values (aka fractions) and are also categorized by size. Again the only real guarantee you have is that sizeof(float) <= sizeof(double) <= sizeof (long double). Floating-point types are stored using a rather peculiar memory model that can be difficult to understand, and that I won't go into, but there is an excellent guide here.
There's a fantastic blog post about C primitives in an Objective-C context over at RyPress. Lots of intro CPS textbooks also have good resources.
Firstly I would like to specify the difference between au unsigned int and an int. Say that you have a very high number, and that you write a loop iterating with an unsigned int:
for(unsigned int i=0; i< N; i++)
{ ... }
If N is a number defined with #define, it may be higher that the maximum value storable with an int instead of an unsigned int. If you overflow i will start again from zero and you'll go in an infinite loop, that's why I prefer to use an int for loops.
The same happens if for mistake you iterate with an int, comparing it to a long. If N is a long you should iterate with a long, but if N is an int you can still safely iterate with a long.
Another pitfail that may occur is when using the shift operator with an integer constant, then assigning it to an int or long. Maybe you also log sizeof(long) and you notice that it returns 8 and you don't care about portability, so you think that you wouldn't lose precision here:
long i= 1 << 34;
Bit instead 1 isn't a long, so it will overflow and when you cast it to a long you have already lost precision. Instead you should type:
long i= 1l << 34;
Newer compilers will warn you about this.
Taken from this question: Converting Long 64-bit Decimal to Binary.
About float and double there is a thing to considerate: they use a mantissa and an exponent to represent the number. It's something like:
value= 2^exponent * mantissa
So the more the exponent is high, the more the floating point number doesn't have an exact representation. It may also happen that a number is too high, so that it will have a so inaccurate representation, that surprisingly if you print it you get a different number:
float f= 9876543219124567;
NSLog("%.0f",f); // On my machine it prints 9876543585124352
If I use a double it prints 9876543219124568, and if I use a long double with the .0Lf format it prints the correct value. Always be careful when using floating points numbers, unexpected things may happen.
For example it may also happen that two floating point numbers have almost the same value, that you expect they have the same value but there is a subtle difference, so that the equality comparison fails. But this has been treated hundreds of times on Stack Overflow, so I will just post this link: What is the most effective way for float and double comparison?.
For the problem of producing a bit-pattern with exactly n set bits, I know of two practical methods, but they both have limitations I'm not happy with.
First, you can enumerate all of the possible word values which have that many bits set in a pre-computed table, and then generate a random index into that table to pick out a possible result. This has the problem that as the output size grows the list of candidate outputs eventually becomes impractically large.
Alternatively, you can pick n non-overlapping bit positions at random (for example, by using a partial Fisher-Yates shuffle) and set those bits only. This approach, however, computes a random state in a much larger space than the number of possible results. For example, it may choose the first and second bits out of three, or it might, separately, choose the second and first bits.
This second approach must consume more bits from the random number source than are strictly required. Since it is choosing n bits in a specific order when their order is unimportant, this means that it is making an arbitrary distinction between n! different ways of producing the same result, and consuming at least floor(log_2(n!)) more bits than are necessary.
Can this be avoided?
There is obviously a third approach of iteratively computing and counting off the legal permutations until a random index is reached, but that's simply a space-for-time trade-off on the first approach, and isn't directly helpful unless there is an efficient way to count off those n permutations.
clarification
The first approach requires picking a single random number between zero and (where w is the output size), as this is the number of possible solutions.
The second approach requires picking n random values between zero and w-1, zero and w-2, etc., and these have a product of , which is times larger than the first approach.
This means that the random number source has been forced to produce bits to distinguish n! different results which are all equivalent. I'd like to know if there's an efficient method to avoid relying on this superfluous randomness. Perhaps by using an algorithm which produces an un-ordered list of bit positions, or by directly computing the nth unique permutation of bits.
Seems like you want a variant of Floyd's algorithm:
Algorithm to select a single, random combination of values?
Should be especially useful in your case, because the containment test is a simple bitmask operation. This will require only k calls to the RNG. In the code below, I assume you have randint(limit) which produces a uniform random from 0 to limit-1, and that you want k bits set in a 32-bit int:
mask = 0;
for (j = 32 - k; j < 32; ++j) {
r = randint(j+1);
b = 1 << r;
if (mask & b) mask |= (1 << j);
else mask |= b;
}
How many bits of entropy you need here depends on how randint() is implemented. If k > 16, set it to 32 - k and negate the result.
Your alternative suggestion of generating a single random number representing one combination among the set (mathematicians would call this a rank of the combination) is simpler if you use colex order rather than lexicographic rank. This code, for example:
for (i = k; i >= 1; --i) {
while ((b = binomial(n, i)) > r) --n;
buf[i-1] = n;
r -= b;
}
will fill the array buf[] with indices from 0 to n-1 for the k-combination at colex rank r. In your case, you'd replace buf[i-1] = n with mask |= (1 << n). The binomial() function is binomial coefficient, which I do with a lookup table (see this). That would make the most efficient use of entropy, but I still think Floyd's algorithm would be a better compromise.
[Expanding my comment:] If you only have a little raw entropy available, then use a PRNG to stretch it further. You only need enough raw entropy to seed a PRNG. Use the PRNG to do the actual shuffle, not the raw entropy. For the next shuffle reseed the PRNG with some more raw entropy. That spreads out the raw entropy and makes less of a demand on your entropy source.
If you know exactly the range of numbers you need out of the PRNG, then you can, carefully, set up your own LCG PRNG to cover the appropriate range while needing the minimum entropy to seed it.
ETA: In C++there is a next_permutation() method. Try using that. See std::next_permutation Implementation Explanation for more.
Is this a theory problem or a practical problem?
You could still do the partial shuffle, but keep track of the order of the ones and forget the zeroes. There are log(k!) bits of unused entropy in their final order for your future consumption.
You could also just use the recurrence (n choose k) = (n-1 choose k-1) + (n-1 choose k) directly. Generate a random number between 0 and (n choose k)-1. Call it r. Iterate over all of the bits from the nth to the first. If we have to set j of the i remaining bits, set the ith if r < (i-1 choose j-1) and clear it, subtracting (i-1 choose j-1), otherwise.
Practically, I wouldn't worry about the couple of words of wasted entropy from the partial shuffle; generating a random 32-bit word with 16 bits set costs somewhere between 64 and 80 bits of entropy, and that's entirely acceptable. The growth rate of the required entropy is asymptotically worse than the theoretical bound, so I'd do something different for really big words.
For really big words, you might generate n independent bits that are 1 with probability k/n. This immediately blows your entropy budget (and then some), but it only uses linearly many bits. The number of set bits is tightly concentrated around k, though. For a further expected linear entropy cost, I can fix it up. This approach has much better memory locality than the partial shuffle approach, so I'd probably prefer it in practice.
I would use solution number 3, generate the i-th permutation.
But do you need to generate the first i-1 ones?
You can do it a bit faster than that with kind of divide and conquer method proposed here: Returning i-th combination of a bit array and maybe you can improve the solution a bit
Background
From the formula you have given - w! / ((w-n)! * n!) it looks like your problem set has to do with the binomial coefficient which deals with calculating the number of unique combinations and not permutations which deals with duplicates in different positions.
You said:
"There is obviously a third approach of iteratively computing and counting off the legal permutations until a random index is reached, but that's simply a space-for-time trade-off on the first approach, and isn't directly helpful unless there is an efficient way to count off those n permutations.
...
This means that the random number source has been forced to produce bits to distinguish n! different results which are all equivalent. I'd like to know if there's an efficient method to avoid relying on this superfluous randomness. Perhaps by using an algorithm which produces an un-ordered list of bit positions, or by directly computing the nth unique permutation of bits."
So, there is a way to efficiently compute the nth unique combination, or rank, from the k-indexes. The k-indexes refers to a unique combination. For example, lets say that the n choose k case of 4 choose 3 is taken. This means that there are a total of 4 numbers that can be selected (0, 1, 2, 3), which is represented by n, and they are taken in groups of 3, which is represented by k. The total number of unique combinations can be calculated as n! / ((k! * (n-k)!). The rank of zero corresponds to the k-index of (2, 1, 0). Rank one is represented by the k-index group of (3, 1, 0), and so forth.
Solution
There is a formula that can be used to very efficiently translate between a k-index group and the corresponding rank without iteration. Likewise, there is a formula for translating between the rank and corresponding k-index group.
I have written a paper on this formula and how it can be seen from Pascal's Triangle. The paper is called Tablizing The Binomial Coeffieicent.
I have written a C# class which is in the public domain that implements the formula described in the paper. It uses very little memory and can be downloaded from the site. It performs the following tasks:
Outputs all the k-indexes in a nice format for any N choose K to a file. The K-indexes can be substituted with more descriptive strings or letters.
Converts the k-index to the proper lexicographic index or rank of an entry in the sorted binomial coefficient table. This technique is much faster than older published techniques that rely on iteration. It does this by using a mathematical property inherent in Pascal's Triangle and is very efficient compared to iterating over the entire set.
Converts the index in a sorted binomial coefficient table to the corresponding k-index. The technique used is also much faster than older iterative solutions.
Uses Mark Dominus method to calculate the binomial coefficient, which is much less likely to overflow and works with larger numbers. This version returns a long value. There is at least one other method that returns an int. Make sure that you use the method that returns a long value.
The class is written in .NET C# and provides a way to manage the objects related to the problem (if any) by using a generic list. The constructor of this class takes a bool value called InitTable that when true will create a generic list to hold the objects to be managed. If this value is false, then it will not create the table. The table does not need to be created in order to use the 4 above methods. Accessor methods are provided to access the table.
There is an associated test class which shows how to use the class and its methods. It has been extensively tested with at least 2 cases and there are no known bugs.
The following tested example code demonstrates how to use the class and will iterate through each unique combination:
public void Test10Choose5()
{
String S;
int Loop;
int N = 10; // Total number of elements in the set.
int K = 5; // Total number of elements in each group.
// Create the bin coeff object required to get all
// the combos for this N choose K combination.
BinCoeff<int> BC = new BinCoeff<int>(N, K, false);
int NumCombos = BinCoeff<int>.GetBinCoeff(N, K);
// The Kindexes array specifies the indexes for a lexigraphic element.
int[] KIndexes = new int[K];
StringBuilder SB = new StringBuilder();
// Loop thru all the combinations for this N choose K case.
for (int Combo = 0; Combo < NumCombos; Combo++)
{
// Get the k-indexes for this combination.
BC.GetKIndexes(Combo, KIndexes);
// Verify that the Kindexes returned can be used to retrive the
// rank or lexigraphic order of the KIndexes in the table.
int Val = BC.GetIndex(true, KIndexes);
if (Val != Combo)
{
S = "Val of " + Val.ToString() + " != Combo Value of " + Combo.ToString();
Console.WriteLine(S);
}
SB.Remove(0, SB.Length);
for (Loop = 0; Loop < K; Loop++)
{
SB.Append(KIndexes[Loop].ToString());
if (Loop < K - 1)
SB.Append(" ");
}
S = "KIndexes = " + SB.ToString();
Console.WriteLine(S);
}
}
So, the way to apply the class to your problem is by considering each bit in the word size as the total number of items. This would be n in the n!/((k! (n - k)!) formula. To obtain k, or the group size, simply count the number of bits set to 1. You would have to create a list or array of the class objects for each possible k, which in this case would be 32. Note that the class does not handle N choose N, N choose 0, or N choose 1 so the code would have to check for those cases and return 1 for both the 32 choose 0 case and 32 choose 32 case. For 32 choose 1, it would need to return 32.
If you need to use values not much larger than 32 choose 16 (the worst case for 32 items - yields 601,080,390 unique combinations), then you can use 32 bit integers, which is how the class is currently implemented. If you need to use 64 bit integers, then you will have to convert the class to use 64 bit longs. The largest value that a long can hold is 18,446,744,073,709,551,616 which is 2 ^ 64. The worst case for n choose k when n is 64 is 64 choose 32. 64 choose 32 is 1,832,624,140,942,590,534 - so a long value will work for all 64 choose k cases. If you need numbers bigger than that, then you will probably want to look into using some sort of big integer class. In C#, the .NET framework has a BigInteger class. If you are working in a different language, it should not be hard to port.
If you are looking for a very good PRNG, one of the fastest, lightweight, and high quality output is the Tiny Mersenne Twister or TinyMT for short . I ported the code over to C++ and C#. it can be found here, along with a link to the original author's C code.
Rather than using a shuffling algorithm like Fisher-Yates, you might consider doing something like the following example instead:
// Get 7 random cards.
ulong Card;
ulong SevenCardHand = 0;
for (int CardLoop = 0; CardLoop < 7; CardLoop++)
{
do
{
// The card has a value of between 0 and 51. So, get a random value and
// left shift it into the proper bit position.
Card = (1UL << RandObj.Next(CardsInDeck));
} while ((SevenCardHand & Card) != 0);
SevenCardHand |= Card;
}
The above code is faster than any shuffling algorithm (at least for obtaining a subset of random cards) since it only works on 7 cards instead of 52. It also packs the cards into individual bits within a single 64 bit word. It makes evaluating poker hands much more efficient as well.
As a side, note, the best binomial coefficient calculator I have found that works with very large numbers (it accurately calculated a case that yielded over 15,000 digits in the result) can be found here.
What is a quick and easy way to 'checksum' an array of floating point numbers, while allowing for a specified small amount of inaccuracy?
e.g. I have two algorithms which should (in theory, with infinite precision) output the same array. But they work differently, and so floating point errors will accumulate differently, though the array lengths should be exactly the same. I'd like a quick and easy way to test if the arrays seem to be the same. I could of course compare the numbers pairwise, and report the maximum error; but one algorithm is in C++ and the other is in Mathematica and I don't want the bother of writing out the numbers to a file or pasting them from one system to another. That's why I want a simple checksum.
I could simply add up all the numbers in the array. If the array length is N, and I can tolerate an error of 0.0001 in each number, then I would check if abs(sum1-sum2)<0.0001*N. But this simplistic 'checksum' is not robust, e.g. to an error of +10 in one entry and -10 in another. (And anyway, probability theory says that the error probably grows like sqrt(N), not like N.) Of course, any checksum is a low-dimensional summary of a chunk of data so it will miss some errors, if not most... but simple checksums are nonetheless useful for finding non-malicious bug-type errors.
Or I could create a two-dimensional checksum, [sum(x[n]), sum(abs(x[n]))]. But is the best I can do, i.e. is there a different function I might use that would be "more orthogonal" to the sum(x[n])? And if I used some arbitrary functions, e.g. [sum(f1(x[n])), sum(f2(x[n]))], then how should my 'raw error tolerance' translate into 'checksum error tolerance'?
I'm programming in C++, but I'm happy to see answers in any language.
i have a feeling that what you want may be possible via something like gray codes. if you could translate your values into gray codes and use some kind of checksum that was able to correct n bits you could detect whether or not the two arrays were the same except for n-1 bits of error, right? (each bit of error means a number is "off by one", where the mapping would be such that this was a variation in the least significant digit).
but the exact details are beyond me - particularly for floating point values.
i don't know if it helps, but what gray codes solve is the problem of pathological rounding. rounding sounds like it will solve the problem - a naive solution might round and then checksum. but simple rounding always has pathological cases - for example, if we use floor, then 0.9999999 and 1 are distinct. a gray code approach seems to address that, since neighbouring values are always single bit away, so a bit-based checksum will accurately reflect "distance".
[update:] more exactly, what you want is a checksum that gives an estimate of the hamming distance between your gray-encoded sequences (and the gray encoded part is easy if you just care about 0.0001 since you can multiple everything by 10000 and use integers).
and it seems like such checksums do exist: Any error-correcting code can be used for error detection. A code with minimum Hamming distance, d, can detect up to d − 1 errors in a code word. Using minimum-distance-based error-correcting codes for error detection can be suitable if a strict limit on the minimum number of errors to be detected is desired.
so, just in case it's not clear:
multiple by minimum error to get integers
convert to gray code equivalent
use an error detecting code with a minimum hamming distance larger than the error you can tolerate.
but i am still not sure that's right. you still get the pathological rounding in the conversion from float to integer. so it seems like you need a minimum hamming distance that is 1 + len(data) (worst case, with a rounding error on each value). is that feasible? probably not for large arrays.
maybe ask again with better tags/description now that a general direction is possible? or just add tags now? we need someone who does this for a living. [i added a couple of tags]
I've spent a while looking for a deterministic answer, and been unable to find one. If there is a good answer, it's likely to require heavy-duty mathematical skills (functional analysis).
I'm pretty sure there is no solution based on "discretize in some cunning way, then apply a discrete checksum", e.g. "discretize into strings of 0/1/?, where ? means wildcard". Any discretization will have the property that two floating-point numbers very close to each other can end up with different discrete codes, and then the discrete checksum won't tell us what we want to know.
However, a very simple randomized scheme should work fine. Generate a pseudorandom string S from the alphabet {+1,-1}, and compute csx=sum(X_i*S_i) and csy=sum(Y_i*S_i), where X and Y are my original arrays of floating point numbers. If we model the errors as independent Normal random variables with mean 0, then it's easy to compute the distribution of csx-csy. We could do this for several strings S, and then do a hypothesis test that the mean error is 0. The number of strings S needed for the test is fixed, it doesn't grow linearly in the size of the arrays, so it satisfies my need for a "low-dimensional summary". This method also gives an estimate of the standard deviation of the error, which may be handy.
Try this:
#include <complex>
#include <cmath>
#include <iostream>
// PARAMETERS
const size_t no_freqs = 3;
const double freqs[no_freqs] = {0.05, 0.16, 0.39}; // (for example)
int main() {
std::complex<double> spectral_amplitude[no_freqs];
for (size_t i = 0; i < no_freqs; ++i) spectral_amplitude[i] = 0.0;
size_t n_data = 0;
{
std::complex<double> datum;
while (std::cin >> datum) {
for (size_t i = 0; i < no_freqs; ++i) {
spectral_amplitude[i] += datum * std::exp(
std::complex<double>(0.0, 1.0) * freqs[i] * double(n_data)
);
}
++n_data;
}
}
std::cout << "Fuzzy checksum:\n";
for (size_t i = 0; i < no_freqs; ++i) {
std::cout << real(spectral_amplitude[i]) << "\n";
std::cout << imag(spectral_amplitude[i]) << "\n";
}
std::cout << "\n";
return 0;
}
It returns just a few, arbitrary points of a Fourier transform of the entire data set. These make a fuzzy checksum, so to speak.
How about computing a standard integer checksum on the data obtained by zeroing the least significant digits of the data, the ones that you don't care about?
There's a common way to store multiple values in one variable, by using a bitmask. For example, if a user has read, write and execute privileges on an item, that can be converted to a single number by saying read = 4 (2^2), write = 2 (2^1), execute = 1 (2^0) and then add them together to get 7.
I use this technique in several web applications, where I'd usually store the variable into a field and give it a type of MEDIUMINT or whatever, depending on the number of different values.
What I'm interested in, is whether or not there is a practical limit to the number of values you can store like this? For example, if the number was over 64, you couldn't use (64 bit) integers any more. If this was the case, what would you use? How would it affect your program logic (ie: could you still use bitwise comparisons)?
I know that once you start getting really large sets of values, a different method would be the optimal solution, but I'm interested in the boundaries of this method.
Off the top of my head, I'd write a set_bit and get_bit function that could take an array of bytes and a bit offset in the array, and use some bit-twiddling to set/get the appropriate bit in the array. Something like this (in C, but hopefully you get the idea):
// sets the n-th bit in |bytes|. num_bytes is the number of bytes in the array
// result is 0 on success, non-zero on failure (offset out-of-bounds)
int set_bit(char* bytes, unsigned long num_bytes, unsigned long offset)
{
// make sure offset is valid
if(offset < 0 || offset > (num_bytes<<3)-1) { return -1; }
//set the right bit
bytes[offset >> 3] |= (1 << (offset & 0x7));
return 0; //success
}
//gets the n-th bit in |bytes|. num_bytes is the number of bytes in the array
// returns (-1) on error, 0 if bit is "off", positive number if "on"
int get_bit(char* bytes, unsigned long num_bytes, unsigned long offset)
{
// make sure offset is valid
if(offset < 0 || offset > (num_bytes<<3)-1) { return -1; }
//get the right bit
return (bytes[offset >> 3] & (1 << (offset & 0x7));
}
I've used bit masks in filesystem code where the bit mask is many times bigger than a machine word. think of it like an "array of booleans";
(journalling masks in flash memory if you want to know)
many compilers know how to do this for you. Adda bit of OO code to have types that operate senibly and then your code starts looking like it's intent, not some bit-banging.
My 2 cents.
With a 64-bit integer, you can store values up to 2^64-1, 64 is only 2^6. So yes, there is a limit, but if you need more than 64-its worth of flags, I'd be very interested to know what they were all doing :)
How many states so you need to potentially think about? If you have 64 potential states, the number of combinations they can exist in is the full size of a 64-bit integer.
If you need to worry about 128 flags, then a pair of bit vectors would suffice (2^64 * 2).
Addition: in Programming Pearls, there is an extended discussion of using a bit array of length 10^7, implemented in integers (for holding used 800 numbers) - it's very fast, and very appropriate for the task described in that chapter.
Some languages ( I believe perl does, not sure ) permit bitwise arithmetic on strings. Giving you a much greater effective range. ( (strlen * 8bit chars ) combinations )
However, I wouldn't use a single value for superimposition of more than one /type/ of data. The basic r/w/x triplet of 3-bit ints would probably be the upper "practical" limit, not for space efficiency reasons, but for practical development reasons.
( Php uses this system to control its error-messages, and I have already found that its a bit over-the-top when you have to define values where php's constants are not resident and you have to generate the integer by hand, and to be honest, if chmod didn't support the 'ugo+rwx' style syntax I'd never want to use it because i can never remember the magic numbers )
The instant you have to crack open a constants table to debug code you know you've gone too far.
Old thread, but it's worth mentioning that there are cases requiring bloated bit masks, e.g., molecular fingerprints, which are often generated as 1024-bit arrays which we have packed in 32 bigint fields (SQL Server not supporting UInt32). Bit wise operations work fine - until your table starts to grow and you realize the sluggishness of separate function calls. The binary data type would work, were it not for T-SQL's ban on bitwise operators having two binary operands.
For example .NET uses array of integers as an internal storage for their BitArray class.
Practically there's no other way around.
That being said, in SQL you will need more than one column (or use the BLOBS) to store all the states.
You tagged this question SQL, so I think you need to consult with the documentation for your database to find the size of an integer. Then subtract one bit for the sign, just to be safe.
Edit: Your comment says you're using MySQL. The documentation for MySQL 5.0 Numeric Types states that the maximum size of a NUMERIC is 64 or 65 digits. That's 212 bits for 64 digits.
Remember that your language of choice has to be able to work with those digits, so you may be limited to a 64-bit integer anyway.
What is the best method for comparing IEEE floats and doubles for equality? I have heard of several methods, but I wanted to see what the community thought.
The best approach I think is to compare ULPs.
bool is_nan(float f)
{
return (*reinterpret_cast<unsigned __int32*>(&f) & 0x7f800000) == 0x7f800000 && (*reinterpret_cast<unsigned __int32*>(&f) & 0x007fffff) != 0;
}
bool is_finite(float f)
{
return (*reinterpret_cast<unsigned __int32*>(&f) & 0x7f800000) != 0x7f800000;
}
// if this symbol is defined, NaNs are never equal to anything (as is normal in IEEE floating point)
// if this symbol is not defined, NaNs are hugely different from regular numbers, but might be equal to each other
#define UNEQUAL_NANS 1
// if this symbol is defined, infinites are never equal to finite numbers (as they're unimaginably greater)
// if this symbol is not defined, infinities are 1 ULP away from +/- FLT_MAX
#define INFINITE_INFINITIES 1
// test whether two IEEE floats are within a specified number of representable values of each other
// This depends on the fact that IEEE floats are properly ordered when treated as signed magnitude integers
bool equal_float(float lhs, float rhs, unsigned __int32 max_ulp_difference)
{
#ifdef UNEQUAL_NANS
if(is_nan(lhs) || is_nan(rhs))
{
return false;
}
#endif
#ifdef INFINITE_INFINITIES
if((is_finite(lhs) && !is_finite(rhs)) || (!is_finite(lhs) && is_finite(rhs)))
{
return false;
}
#endif
signed __int32 left(*reinterpret_cast<signed __int32*>(&lhs));
// transform signed magnitude ints into 2s complement signed ints
if(left < 0)
{
left = 0x80000000 - left;
}
signed __int32 right(*reinterpret_cast<signed __int32*>(&rhs));
// transform signed magnitude ints into 2s complement signed ints
if(right < 0)
{
right = 0x80000000 - right;
}
if(static_cast<unsigned __int32>(std::abs(left - right)) <= max_ulp_difference)
{
return true;
}
return false;
}
A similar technique can be used for doubles. The trick is to convert the floats so that they're ordered (as if integers) and then just see how different they are.
I have no idea why this damn thing is screwing up my underscores. Edit: Oh, perhaps that is just an artefact of the preview. That's OK then.
The current version I am using is this
bool is_equals(float A, float B,
float maxRelativeError, float maxAbsoluteError)
{
if (fabs(A - B) < maxAbsoluteError)
return true;
float relativeError;
if (fabs(B) > fabs(A))
relativeError = fabs((A - B) / B);
else
relativeError = fabs((A - B) / A);
if (relativeError <= maxRelativeError)
return true;
return false;
}
This seems to take care of most problems by combining relative and absolute error tolerance. Is the ULP approach better? If so, why?
#DrPizza: I am no performance guru but I would expect fixed point operations to be quicker than floating point operations (in most cases).
It rather depends on what you are doing with them. A fixed-point type with the same range as an IEEE float would be many many times slower (and many times larger).
Things suitable for floats:
3D graphics, physics/engineering, simulation, climate simulation....
In numerical software you often want to test whether two floating point numbers are exactly equal. LAPACK is full of examples for such cases. Sure, the most common case is where you want to test whether a floating point number equals "Zero", "One", "Two", "Half". If anyone is interested I can pick some algorithms and go more into detail.
Also in BLAS you often want to check whether a floating point number is exactly Zero or One. For example, the routine dgemv can compute operations of the form
y = beta*y + alpha*A*x
y = beta*y + alpha*A^T*x
y = beta*y + alpha*A^H*x
So if beta equals One you have an "plus assignment" and for beta equals Zero a "simple assignment". So you certainly can cut the computational cost if you give these (common) cases a special treatment.
Sure, you could design the BLAS routines in such a way that you can avoid exact comparisons (e.g. using some flags). However, the LAPACK is full of examples where it is not possible.
P.S.:
There are certainly many cases where you don't want check for "is exactly equal". For many people this even might be the only case they ever have to deal with. All I want to point out is that there are other cases too.
Although LAPACK is written in Fortran the logic is the same if you are using other programming languages for numerical software.
Oh dear lord please don't interpret the float bits as ints unless you're running on a P6 or earlier.
Even if it causes it to copy from vector registers to integer registers via memory, and even if it stalls the pipeline, it's the best way to do it that I've come across, insofar as it provides the most robust comparisons even in the face of floating point errors.
i.e. it is a price worth paying.
This seems to take care of most problems by combining relative and absolute error tolerance. Is the ULP approach better? If so, why?
ULPs are a direct measure of the "distance" between two floating point numbers. This means that they don't require you to conjure up the relative and absolute error values, nor do you have to make sure to get those values "about right". With ULPs, you can express directly how close you want the numbers to be, and the same threshold works just as well for small values as for large ones.
If you have floating point errors you have even more problems than this. Although I guess that is up to personal perspective.
Even if we do the numeric analysis to minimize accumulation of error, we can't eliminate it and we can be left with results that ought to be identical (if we were calculating with reals) but differ (because we cannot calculate with reals).
If you are looking for two floats to be equal, then they should be identically equal in my opinion. If you are facing a floating point rounding problem, perhaps a fixed point representation would suit your problem better.
If you are looking for two floats to be equal, then they should be identically equal in my opinion. If you are facing a floating point rounding problem, perhaps a fixed point representation would suit your problem better.
Perhaps we cannot afford the loss of range or performance that such an approach would inflict.
#DrPizza: I am no performance guru but I would expect fixed point operations to be quicker than floating point operations (in most cases).
#Craig H: Sure. I'm totally okay with it printing that. If a or b store money then they should be represented in fixed point. I'm struggling to think of a real world example where such logic ought to be allied to floats. Things suitable for floats:
weights
ranks
distances
real world values (like from a ADC)
For all these things, either you much then numbers and simply present the results to the user for human interpretation, or you make a comparative statement (even if such a statement is, "this thing is within 0.001 of this other thing"). A comparative statement like mine is only useful in the context of the algorithm: the "within 0.001" part depends on what physical question you're asking. That my 0.02. Or should I say 2/100ths?
It rather depends on what you are
doing with them. A fixed-point type
with the same range as an IEEE float
would be many many times slower (and
many times larger).
Okay, but if I want a infinitesimally small bit-resolution then it's back to my original point: == and != have no meaning in the context of such a problem.
An int lets me express ~10^9 values (regardless of the range) which seems like enough for any situation where I would care about two of them being equal. And if that's not enough, use a 64-bit OS and you've got about 10^19 distinct values.
I can express values a range of 0 to 10^200 (for example) in an int, it is just the bit-resolution that suffers (resolution would be greater than 1, but, again, no application has that sort of range as well as that sort of resolution).
To summarize, I think in all cases one either is representing a continuum of values, in which case != and == are irrelevant, or one is representing a fixed set of values, which can be mapped to an int (or a another fixed-precision type).
An int lets me express ~10^9 values
(regardless of the range) which seems
like enough for any situation where I
would care about two of them being
equal. And if that's not enough, use a
64-bit OS and you've got about 10^19
distinct values.
I have actually hit that limit... I was trying to juggle times in ps and time in clock cycles in a simulation where you easily hit 10^10 cycles. No matter what I did I very quickly overflowed the puny range of 64-bit integers... 10^19 is not as much as you think it is, gimme 128 bits computing now!
Floats allowed me to get a solution to the mathematical issues, as the values overflowed with lots zeros at the low end. So you basically had a decimal point floating aronud in the number with no loss of precision (I could like with the more limited distinct number of values allowed in the mantissa of a float compared to a 64-bit int, but desperately needed th range!).
And then things converted back to integers to compare etc.
Annoying, and in the end I scrapped the entire attempt and just relied on floats and < and > to get the work done. Not perfect, but works for the use case envisioned.
If you are looking for two floats to be equal, then they should be identically equal in my opinion. If you are facing a floating point rounding problem, perhaps a fixed point representation would suit your problem better.
Perhaps I should explain the problem better. In C++, the following code:
#include <iostream>
using namespace std;
int main()
{
float a = 1.0;
float b = 0.0;
for(int i=0;i<10;++i)
{
b+=0.1;
}
if(a != b)
{
cout << "Something is wrong" << endl;
}
return 1;
}
prints the phrase "Something is wrong". Are you saying that it should?
Oh dear lord please don't interpret the float bits as ints unless you're running on a P6 or earlier.
it's the best way to do it that I've come across, insofar as it provides the most robust comparisons even in the face of floating point errors.
If you have floating point errors you have even more problems than this. Although I guess that is up to personal perspective.