“Methods of computing square roots”的意思、由来-开放百科全书

In numerical analysis, a branch of mathematics, there are several square root algorithms or methods of computing the principal square root of a non-negative real number. For the square roots of a negative or complex number, see below.

Finding

is the same as solving the equation

for a positive

. Therefore, any general numerical root-finding algorithm can be used. Newton's method, for example, reduces in this case to the so-called Babylonian method:

These methods generally yield approximate results, but can be made arbitrarily precise by increasing the number of calculation steps.

Rough estimation

Many square root algorithms require an initial seed value. If the initial seed value is far away from the actual square root, the algorithm will be slowed down. It is therefore useful to have a rough estimate, which may be very inaccurate but easy to calculate. With

expressed in scientific notation as

where

and n is an integer, the square root

can be estimated as

The factors two and six are used because they approximate the geometric means of the lowest and highest possible values with the given number of digits:

and

When working in the binary numeral system (as computers do internally), by expressing

where

, the square root

can be estimated as

, since the geometric mean of the lowest and highest possible values is

These approximations are useful to find better seeds for iterative algorithms, which results in faster convergence.

Perhaps the first algorithm used for approximating

is known as the Babylonian method, despite there being no direct evidence, beyond informed conjecture, that the eponymous Babylonian mathematicians employed exactly this method.^[1] The method is also known as Heron's method, after the first-century Greek mathematician Hero of Alexandria who gave the first explicit description of the method in his AD 60 work Metrica.^[2] The basic idea is that if {{mvar|x}} is an overestimate to the square root of a non-negative real number {{mvar|S}} then {{math|{{sfrac|S|x}}}} will be an underestimate, or vice versa, and so the average of these two numbers may reasonably be expected to provide a better approximation (though the formal proof of that assertion depends on the inequality of arithmetic and geometric means that shows this average is always an overestimate of the square root, as noted in the article on square roots, thus assuring convergence).

More precisely, if {{mvar|x}} is our initial guess of

and {{mvar|e}} is the error in our estimate such that {{math|1=S = (x+ e)²}}, then we can expand the binomial and solve for

Since the computed error was not exact, this becomes our next best guess. The process of updating is iterated until desired accuracy is obtained. This is a quadratically convergent algorithm, which means that the number of correct digits of the approximation roughly doubles with each iteration. It proceeds as follows:

This algorithm works equally well in the {{mvar|p}}-adic numbers, but cannot be used to identify real square roots with {{mvar|p}}-adic square roots; one can, for example, construct a sequence of rational numbers by this method that converges to +3 in the reals, but to −3 in the 2-adics.

Example

To calculate {{math|{{sqrt|S}}}}, where {{mvar|S}} = 125348, to six significant figures, use the rough estimation method above to get

Convergence

Suppose that x₀ > 0 and S > 0. Then for any natural number n, x_n > 0. Let the relative error in x_n be defined by

Worst case for convergence

If using the rough estimate above with the Babylonian method, then the least accurate cases in ascending order are as follows:

Rounding errors will slow the convergence. It is recommended to keep at least one extra digit beyond the desired accuracy of the {{mvar|x_n}} being calculated to minimize round off error.

Bakhshali method

This method for finding an approximation to a square root was described in an ancient Indian mathematical manuscript called the Bakhshali manuscript. It is equivalent to two iterations of the Babylonian method beginning with x₀. Thus, the algorithm is quartically convergent, which means that the number of correct digits of the approximation roughly quadruples with each iteration.^[3] The original presentation, using modern notation, is as follows: To calculate

, let x₀² be the initial approximation to S. Then, successively iterate as:

Let x₀ = N be an integer which is the nearest perfect square to S. Also, let the difference d = S - N², then the first iteration can be written as:

Example

Using the same example as given in Babylonian method, let

Then, the first iterations gives

Digit-by-digit calculation

This is a method to find each digit of the square root in a sequence. It is slower than the Babylonian method, but it has several advantages:

Napier's bones include an aid for the execution of this algorithm. The shifting nth root algorithm is a generalization of this method.

Basic principle

First, let's consider the simplest possible case of finding the square root of a number Z, that is the square of a 2 digit number XY, where X is the tens digit and Y is the units digit. Specifically:

Now using the Digit-by-Digit algorithm, we first determine the value of X. X is the largest digit such that X² is less or equal to Z from which we removed the 2 rightmost digits.

In the next iteration, we pair the digits, multiply X by 2, and place it in the tenth's place while we try to figure out what the value of Y is.

Since this is a simple case where the answer is a perfect square root XY, the algorithm stops here.

The same idea can be extended to any arbitrary square root computation next. Suppose we are able to find the square root of N by expressing it as a sum of n positive numbers such that

This expression allows us to find the square root by sequentially guessing the values of

s. Suppose that the numbers

have already been guessed, then the m-th term of the right-hand-side of above summation is given by

where

is the approximate square root found so far. Now each new guess

should satisfy the recursion

such that

for all

with initialization

When

the exact square root has been found; if not, then the sum of

s gives a suitable approximation of the square root, with

being the approximation error.

where

are place holders and the coefficients

. At any m-th stage of the square root calculation, the approximate root found so far,

and the summation term

are given by

Here since the place value of

is an even power of 10, we only need to work with the pair of most significant digits of the remaining term

at any m-th stage. The section below codifies this procedure.

It is obvious that a similar method can be used to compute the square root in number systems other than the decimal number system. For instance, finding the digit-by-digit square root in the binary number system is quite efficient since the value of

is searched from a smaller set of binary digits {0,1}. This makes the computation faster since at each stage the value of

is either

for

. The fact that we have only two possible options for

also makes the process of deciding the value of

at m-th stage of calculation easier. This is because we only need to check if

for

If this condition is satisfied, then we take

; if not then

Also, the fact that multiplication by 2 is done by left bit-shifts helps in the computation.

Decimal (base 10)

Write the original number in decimal form. The numbers are written similar to the long division algorithm, and, as in long division, the root will be written on the line above. Now separate the digits into pairs, starting from the decimal point and going both left and right. The decimal point of the root will be above the decimal point of the square. One digit of the root will appear above each pair of digits of the square.

Beginning with the left-most pair of digits, do the following procedure for each pair:

Examples

Binary numeral system (base 2)

Inherent to digit-by-digit algorithms is a search and test step: find a digit,

, when added to the right of a current solution

, such that

, where

is the value for which a root is desired. Expanding:

. The current value of

—or, usually, the remainder—can be incrementally updated efficiently when working in binary, as the value of

will have a single bit set (a power of 2), and the operations needed to compute

and

can be replaced with faster bit shift operations.

Example

Here we obtain the square root of 81, which when converted into binary gives 1010001. The numbers in the left column gives the option between that number or zero to be used for subtraction at that stage of computation. The final answer is 1001, which in decimal is 9.

Using the notation above, the variable "bit" corresponds to

which is

, the variable "res" is equal to

, and the variable "num" is equal to the current

which is the difference of the number we want the square root of and the square of our current approximation with all bits set up to

. Thus in the first loop, we want to find the highest power of 4 in "bit" to find the highest power of 2 in

. In the second loop, if num is greater than res + bit, then

is greater than

and we can subtract it. The next line, we want to add

which means we want to add

so we want res = res + bit<<1. Then update

inside res which involves dividing by 2 or another shift to the right. Combining these 2 into one line leads to res = res>>1 + bit. If

isn't greater than

then we just update

inside res and divide it by 2. Then we update

in bit by dividing it by 4. The final iteration of the 2nd loop has bit equal to 1 and will cause update of

to run one extra time removing the factor of 2 from res making it our integer approximation of the root.

Faster algorithms, in binary and decimal or any other base, can be realized by using lookup tables—in effect trading more storage space for reduced run time.^[5]

Exponential identity

Pocket calculators typically implement good routines to compute the exponential function and the natural logarithm, and then compute the square root of S using the identity found using the properties of logarithms (

) and exponentials (

The denominator in the fraction corresponds to the nth root. In the case above the denominator is 2, hence the equation specifies that the square root is to be found. The same identity is used when computing square roots with logarithm tables or slide rules.

Vedic duplex method for extracting a square root

The Vedic duplex method from the book 'Vedic Mathematics' is a variant of the digit-by-digit method for calculating the square root.^[6] The duplex is the square of the central digit plus double the cross-product of digits equidistant from the center. The duplex is computed from the quotient digits (square root digits) computed thus far, but after the initial digits. The duplex is subtracted from the dividend digit prior to the second subtraction for the product of the quotient digit times the divisor digit. For perfect squares the duplex and the dividend will get smaller and reach zero after a few steps. For non-perfect squares the decimal value of the square root can be calculated to any precision desired. However, as the decimal places proliferate, the duplex adjustment gets larger and longer to calculate. The duplex method follows the Vedic ideal for an algorithm, one-line, mental calculation. It is flexible in choosing the first digit group and the divisor. Small divisors are to be avoided by starting with a larger initial group.

Basic principle

We proceed as with the digit-by-digit calculation by assuming that we want to express a number N as a square of the sum of n positive numbers as

Thus, we can re-express the above identity in terms of the divisor and the duplexes as

such that

for all

, with initialization

When

the algorithm terminates and the sum of

s give the square root. The method is more similar to long division where

is the dividend and

is the remainder.

where

, then the initiation

and the divisor will be

. Also the product at any m-th stage will be

and the duplexes will be

, where

are the duplexes of the sequence

. At any m-th stage, we see that the place value of the duplex

is one less than the product

. Thus, in actual calculations it is customary to subtract the duplex value of the m-th stage at (m+1)-th stage. Also, unlike the previous digit-by-digit square root calculation, where at any given m-th stage, the calculation is done by taking the most significant pair of digits of the remaining term

, the duplex method uses only a single most significant digit of

In other words, to calculate the duplex of a number, double the product of each pair of equidistant digits plus the square of the center digit (of the digits to the right of the colon).

In a square root calculation the quotient digit set increases incrementally for each step.

Example

Consider the perfect square 2809 = 53². Use the duplex method to find the square root of 2,809.

A two-variable iterative method

This method is applicable for finding the square root of

and converges best for

This, however, is no real limitation for a computer based calculation, as in base 2 floating point and fixed point representations, it is trivial to multiply

by an integer power of 4, and therefore

by the corresponding power of 2, by changing the exponent or by shifting, respectively. Therefore,

can be moved to the range

. Moreover, the following method does not employ general divisions, but only additions, subtractions, multiplications, and divisions by powers of two, which are again trivial to implement. A disadvantage of the method is that numerical errors accumulate, in contrast to single variable iterative methods such as the Babylonian one.

The proof of the method is rather easy. First, rewrite the iterative definition of

and therefore the convergence of

to the desired result

is ensured by the convergence of

to 0, which in turn follows from

This method was developed around 1950 by M. V. Wilkes, D. J. Wheeler and S. Gill^[7] for use on EDSAC, one of the first electronic computers.^[8] The method was later generalized, allowing the computation of non-square roots.^[9]

Iterative methods for reciprocal square roots

The following are iterative methods for finding the reciprocal square root of S which is

. Once it has been found, find

by simple multiplication:

. These iterations involve only multiplication, and not division. They are therefore faster than the Babylonian method. However, they are not stable. If the initial value is not close to the reciprocal square root, the iterations will diverge away from it rather than converge to it. It can therefore be advantageous to perform an iteration of the Babylonian method on a rough estimate before starting to apply these methods.

Goldschmidt’s algorithm

Goldschmidt's algorithm finds

faster than Newton-Raphson iteration on a computer with a fused multiply–add instruction and either a pipelined floating point unit or two independent floating-point units.^[10]

until

is sufficiently close to 1, or a fixed number of iterations. The iterations converge to

Note that it is possible to omit either

and

from the computation, and if both are desired then

may be used at the end rather than computing it through in each iteration.

until

is sufficiently close to 0, or a fixed number of iterations. This converges to

Taylor series

If N is an approximation to

, a better approximation can be found by using the Taylor series of the square root function:

As an iterative method, the order of convergence is equal to the number of terms used. With two terms, it is identical to the Babylonian method. With three terms, each iteration takes almost as many operations as the Bakhshali approximation, but converges more slowly.{{Citation needed|date=September 2017}} Therefore, this is not a particularly efficient way of calculation. To maximize the rate of convergence, choose N so that

is as small as possible.

CORDIC

A completely different method for computing the square root is based on the CORDIC algorithm, which uses only very simple operations (addition, subtraction, with bitshift and table lookup to implement multiplication). The square root of S may be obtained as the output

using the hyperbolic coordinate system in vectoring mode, with the following initialization:^[11]

Continued fraction expansion

Quadratic irrationals (numbers of the form

, where a, b and c are integers), and in particular, square roots of integers, have periodic continued fractions. Sometimes what is desired is finding not the numerical value of a square root, but rather its continued fraction expansion, and hence its rational approximation. Let S be the positive number for which we are required to find the square root. Then assuming a to be a number that serves as an initial guess and r to be the remainder term, we can write

Since we have

, we can express the square root of S as

By applying this expression for

to the denominator term of the fraction, we have

Proceeding this way, we get a generalized continued fraction for the square root as

Taking the first three denominators give the rational approximation of {{sqrt|2}} as [1;2,2,2] = 17/12 = 1.41667, correct up to first three decimal places. Taking the first five denominators gives the rational approximation to {{sqrt|2}} as [1;2,2,2,2,2] = 99/70 = 1.4142857, correct up to first five decimal places. Taking more denominators give better approximations.

As another example, for the square root of 3, we can select a = 2 and r = -1, giving us

The first three denominators gives {{sqrt|3}} as 1.73214, correct up to the first four decimal places. Note that it is not necessary to choose an integer valued a. For instance, we can take a = {{sqrt|2}} and r = 1, such that

Algorithm

The following iterative algorithm^[12] can be used to obtain the continued fraction expansion in canonical form (S is any natural number that is not a perfect square):

The algorithm terminates when this triplet is the same as one encountered before.

The algorithm can also terminate on a_i when a_i = 2 a₀,^[13] which is easier to implement.

The expansion will repeat from then on. The sequence [a₀; a₁, a₂, a₃, ...] is the continued fraction expansion:

Example, square root of 114 as a continued fraction

Begin with m₀ = 0; d₀ = 1; and a₀ = 10 (10² = 100 and 11² = 121 > 114 so 10 chosen).

Generalized continued fraction

A more rapid method is to evaluate its generalized continued fraction. From the formula derived there:

and the fact that 114 is 2/3 of the way between 10²=100 and 11²=121 results in

which is simply the aforementioned [10;1,2, 10,2,1, 20,1,2, 10,2,1, 20,1,2, ...] evaluated at every third term. Combining pairs of fractions produces

which is now [10;1,2, 10,2,1,20,1,2, 10,2,1,20,1,2, ...] evaluated at the third term and every six terms thereafter.

Using Pell's equation

Pell's equation (also known as Brahmagupta equation since he was the first to give a solution to this particular equation) and its variants yield a method for efficiently finding continued fraction convergents of square roots of integers. However, it can be complicated to execute, and usually not every convergent is generated. The ideas behind the method are as follows:

Approximations that depend on the floating point representation

A number is represented in a floating point format as

which is also called scientific notation. Its square root is

and similar formulae would apply for cube roots and logarithms. On the face of it, this is no improvement in simplicity, but suppose that only an approximation is required: then just

is good to an order of magnitude. Next, recognise that some powers, p, will be odd, thus for 3141.59 = 3.14159 × 10³ rather than deal with fractional powers of the base, multiply the mantissa by the base and subtract one from the power to make it even. The adjusted representation will become the equivalent of 31.4159 × 10² so that the square root will be {{radic|31.4159}} × 10.

If the integer part of the adjusted mantissa is taken, there can only be the values 1 to 99, and that could be used as an index into a table of 99 pre-computed square roots to complete the estimate. A computer using base sixteen would require a larger table, but one using base two would require only three entries: the possible bits of the integer part of the adjusted mantissa are 01 (the power being even so there was no shift, remembering that a normalised floating point number always has a non-zero high-order digit) or if the power was odd, 10 or 11, these being the first two bits of the original mantissa. Thus, 6.25 = 110.01 in binary, normalised to 1.1001 × 2² an even power so the paired bits of the mantissa are 01, while .625 = 0.101 in binary normalises to 1.01 × 2⁻¹ an odd power so the adjustment is to 10.1 × 2⁻² and the paired bits are 10. Notice that the low order bit of the power is echoed in the high order bit of the pairwise mantissa. An even power has its low-order bit zero and the adjusted mantissa will start with 0, whereas for an odd power that bit is one and the adjusted mantissa will start with 1. Thus, when the power is halved, it is as if its low order bit is shifted out to become the first bit of the pairwise mantissa.

A table with only three entries could be enlarged by incorporating additional bits of the mantissa. However, with computers, rather than calculate an interpolation into a table, it is often better to find some simpler calculation giving equivalent results. Everything now depends on the exact details of the format of the representation, plus what operations are available to access and manipulate the parts of the number. For example, Fortran offers an EXPONENT(x) function to obtain the power. Effort expended in devising a good initial approximation is to be recouped by thereby avoiding the additional iterations of the refinement process that would have been needed for a poor approximation. Since these are few (one iteration requires a divide, an add, and a halving) the constraint is severe.

Many computers follow the IEEE (or sufficiently similar) representation, and a very rapid approximation to the square root can be obtained for starting Newton's method. The technique that follows is based on the fact that the floating point format (in base two) approximates the base-2 logarithm. That is

So for a 32-bit single precision floating point number in IEEE format (where notably, the power has a bias of 127 added for the represented form) you can get the approximate logarithm by interpreting its binary representation as a 32-bit integer, scaling it by

, and removing a bias of 127, i.e.

For example, 1.0 is represented by a hexadecimal number 0x3F800000, which would represent

if taken as an integer. Using the formula above you get

, as expected from

. In a similar fashion you get 0.5 from 1.5 (0x3FC00000).

To get the square root, divide the logarithm by 2 and convert the value back. The following program demonstrates the idea. Note that the exponent's lowest bit is intentionally allowed to propagate into the mantissa. One way to justify the steps in this program is to assume

is the exponent bias and

is the number of explicitly stored bits in the mantissa and then show that

The three mathematical operations forming the core of the above function can be expressed in a single line. An additional adjustment can be added to reduce the maximum relative error. So, the three operations, not including the cast, can be rewritten as

where a is a bias for adjusting the approximation errors. For example, with a = 0 the results are accurate for even powers of 2 (e.g., 1.0), but for other numbers the results will be slightly too big (e.g.,1.5 for 2.0 instead of 1.414... with 6% error). With a = -0x4B0D2, the maximum relative error is minimized to ±3.5%.

If the approximation is to be used for an initial guess for Newton's method to the equation

, then the reciprocal form shown in the following section is preferred.

Reciprocal of the square root

A variant of the above routine is included below, which can be used to compute the reciprocal of the square root, i.e.,

instead, was written by Greg Walsh. The integer-shift approximation produced a relative error of less than 4%, and the error dropped further to 0.15% with one iteration of Newton's method on the following line.^[14] In computer graphics it is a very efficient way to normalize a vector.

Some VLSI hardware implements inverse square root using a second degree polynomial estimation followed by a Goldschmidt iteration.^[15]

Negative or complex square

If S = a+bi where a and b are real and b ≠ 0, then its principal square root is

is the modulus of S. The principal square root of a complex number is defined to be the root with the non-negative real part.

Rough estimation

Example

Convergence

Worst case for convergence

Bakhshali method

Example

Digit-by-digit calculation

Basic principle

Decimal (base 10)

Examples

Binary numeral system (base 2)

Example

Exponential identity

Vedic duplex method for extracting a square root

Basic principle

Example

A two-variable iterative method

Iterative methods for reciprocal square roots

Goldschmidt’s algorithm

Taylor series

CORDIC

Continued fraction expansion

Algorithm

Example, square root of 114 as a continued fraction

Generalized continued fraction

Using Pell's equation

Approximations that depend on the floating point representation

Reciprocal of the square root

Negative or complex square

See also

Notes

External links