问题
What is a very efficient way of determining how many digits there are in an integer in C++?
回答1:
Well, the most efficient way, presuming you know the size of the integer, would be a lookup. Should be faster than the much shorter logarithm based approach. If you don't care about counting the '-', remove the + 1.
// generic solution
template <class T>
int numDigits(T number)
{
int digits = 0;
if (number < 0) digits = 1; // remove this line if '-' counts as a digit
while (number) {
number /= 10;
digits++;
}
return digits;
}
// partial specialization optimization for 32-bit numbers
template<>
int numDigits(int32_t x)
{
if (x == MIN_INT) return 10 + 1;
if (x < 0) return numDigits(-x) + 1;
if (x >= 10000) {
if (x >= 10000000) {
if (x >= 100000000) {
if (x >= 1000000000)
return 10;
return 9;
}
return 8;
}
if (x >= 100000) {
if (x >= 1000000)
return 7;
return 6;
}
return 5;
}
if (x >= 100) {
if (x >= 1000)
return 4;
return 3;
}
if (x >= 10)
return 2;
return 1;
}
// partial-specialization optimization for 8-bit numbers
template <>
int numDigits(char n)
{
// if you have the time, replace this with a static initialization to avoid
// the initial overhead & unnecessary branch
static char x[256] = {0};
if (x[0] == 0) {
for (char c = 1; c != 0; c++)
x[c] = numDigits((int32_t)c);
x[0] = 1;
}
return x[n];
}
回答2:
The simplest way is to do:
unsigned GetNumberOfDigits (unsigned i)
{
return i > 0 ? (int) log10 ((double) i) + 1 : 1;
}
log10 is defined in <cmath> or <math.h>. You'd need to profile this to see if it's faster than any of the others posted here. I'm not sure how robust this is with regards to float point precision. Also, the argument is unsigned as negative values and log don't really mix.
回答3:
Perhaps I misunderstood the question but doesn't this do it?
int NumDigits(int x)
{
x = abs(x);
return (x < 10 ? 1 :
(x < 100 ? 2 :
(x < 1000 ? 3 :
(x < 10000 ? 4 :
(x < 100000 ? 5 :
(x < 1000000 ? 6 :
(x < 10000000 ? 7 :
(x < 100000000 ? 8 :
(x < 1000000000 ? 9 :
10)))))))));
}
回答4:
int digits = 0; while (number != 0) { number /= 10; digits++; }
Note: "0" will have 0 digits! If you need 0 to appear to have 1 digit, use:
int digits = 0; do { number /= 10; digits++; } while (number != 0);
(Thanks Kevin Fegan)
In the end, use a profiler to know which of all the answers here will be faster on your machine...
回答5:
Practical joke: This is the most efficient way (number of digits is calculated at compile-time):
template <unsigned long long N, size_t base=10>
struct numberlength
{
enum { value = 1 + numberlength<N/base, base>::value };
};
template <size_t base>
struct numberlength<0, base>
{
enum { value = 0 };
};
May be useful to determine the width required for number field in formatting, input elements etc.
回答6:
See Bit Twiddling Hacks for a much shorter version of the answer you accepted. It also has the benefit of finding the answer sooner if your input is normally distributed, by checking the big constants first. (v >= 1000000000) catches 76% of the values, so checking that first will on average be faster.
回答7:
convert to string and then use built-in functions
unsigned int i;
cout<< to_string(i).length()<<endl;
回答8:
A previous poster suggested a loop that divides by 10. Since multiplies on modern machines are a lot faster, I'd recommend the following code instead:
int digits = 1, pten=10; while ( pten <= number ) { digits++; pten*=10; }
回答9:
The ppc architecture has a bit counting instruction. With that, you can determine the log base 2 of a positive integer in a single instruction. For example, 32 bit would be:
#define log_2_32_ppc(x) (31-__cntlzw(x))
If you can handle a small margin of error on large values you can convert that to log base 10 with another few instructions:
#define log_10_estimate_32_ppc(x) (9-(((__cntlzw(x)*1233)+1545)>>12))
This is platform specific and slightly inaccurate, but also involves no branches, division or conversion to floating point. All depends on what you need.
I only know the ppc instructions off hand, but other architectures should have similar instructions.
回答10:
int x = 1000;
int numberOfDigits = x ? static_cast<int>(log10(abs(x))) + 1 : 1;
回答11:
#include <iostream>
#include <math.h>
using namespace std;
int main()
{
double num;
int result;
cout<<"Enter a number to find the number of digits, not including decimal places: ";
cin>>num;
result = ((num<=1)? 1 : log10(num)+1);
cout<<"Number of digits "<<result<<endl;
return 0;
}
This is probably the simplest way of solving your problem, assuming you only care about digits before the decimal and assuming anything less than 10 is just 1 digit.
回答12:
#include <stdint.h> // uint32_t [available since C99]
/// Determine the number of digits for a 32 bit integer.
/// - Uses at most 4 comparisons.
/// - (cX) 2014 adolfo.dimare@gmail.com
/// - \see http://stackoverflow.com/questions/1489830/#27669966
/** #d == Number length vs Number of comparisons == #c
\code
#d | #c #d | #c
---+--- ---+---
10 | 4 5 | 4
9 | 4 4 | 4
8 | 3 3 | 3
7 | 3 2 | 3
6 | 3 1 | 3
\endcode
*/
unsigned NumDigits32bs(uint32_t x) {
return // Num-># Digits->[0-9] 32->bits bs->Binary Search
( x >= 100000u // [6-10] [1-5]
? // [6-10]
( x >= 10000000u // [8-10] [6-7]
? // [8-10]
( x >= 100000000u // [9-10] [8]
? // [9-10]
( x >= 1000000000u // [10] [9]
? 10
: 9
)
: 8
)
: // [6-7]
( x >= 1000000u // [7] [6]
? 7
: 6
)
)
: // [1-5]
( x >= 100u // [3-5] [1-2]
? // [3-5]
( x >= 1000u // [4-5] [3]
? // [4-5]
( x >= 10000u // [5] [4]
? 5
: 4
)
: 3
)
: // [1-2]
( x >= 10u // [2] [1]
? 2
: 1
)
)
);
}
回答13:
I like Ira Baxter's answer. Here is a template variant that handles the various sizes and deals with the maximum integer values (updated to hoist the upper bound check out of the loop):
#include <boost/integer_traits.hpp>
template<typename T> T max_decimal()
{
T t = 1;
for (unsigned i = boost::integer_traits<T>::digits10; i; --i)
t *= 10;
return t;
}
template<typename T>
unsigned digits(T v)
{
if (v < 0) v = -v;
if (max_decimal<T>() <= v)
return boost::integer_traits<T>::digits10 + 1;
unsigned digits = 1;
T boundary = 10;
while (boundary <= v) {
boundary *= 10;
++digits;
}
return digits;
}
To actually get the improved performance from hoisting the additional test out of the loop, you need to specialise max_decimal() to return constants for each type on your platform. A sufficiently magic compiler could optimise the call to max_decimal() to a constant, but specialisation is better with most compilers today. As it stands, this version is probably slower because max_decimal costs more than the tests removed from the loop.
I'll leave all that as an exercise for the reader.
回答14:
/// Determine the number of digits for a 64 bit integer.
/// - Uses at most 5 comparisons.
/// - (cX) 2014 adolfo.dimare@gmail.com
/// - \see http://stackoverflow.com/questions/1489830/#27670035
/** #d == Number length vs Number of comparisons == #c
\code
#d | #c #d | #c #d | #c #d | #c
---+--- ---+--- ---+--- ---+---
20 | 5 15 | 5 10 | 5 5 | 5
19 | 5 14 | 5 9 | 5 4 | 5
18 | 4 13 | 4 8 | 4 3 | 4
17 | 4 12 | 4 7 | 4 2 | 4
16 | 4 11 | 4 6 | 4 1 | 4
\endcode
*/
unsigned NumDigits64bs(uint64_t x) {
return // Num-># Digits->[0-9] 64->bits bs->Binary Search
( x >= 10000000000ul // [11-20] [1-10]
?
( x >= 1000000000000000ul // [16-20] [11-15]
? // [16-20]
( x >= 100000000000000000ul // [18-20] [16-17]
? // [18-20]
( x >= 1000000000000000000ul // [19-20] [18]
? // [19-20]
( x >= 10000000000000000000ul // [20] [19]
? 20
: 19
)
: 18
)
: // [16-17]
( x >= 10000000000000000ul // [17] [16]
? 17
: 16
)
)
: // [11-15]
( x >= 1000000000000ul // [13-15] [11-12]
? // [13-15]
( x >= 10000000000000ul // [14-15] [13]
? // [14-15]
( x >= 100000000000000ul // [15] [14]
? 15
: 14
)
: 13
)
: // [11-12]
( x >= 100000000000ul // [12] [11]
? 12
: 11
)
)
)
: // [1-10]
( x >= 100000ul // [6-10] [1-5]
? // [6-10]
( x >= 10000000ul // [8-10] [6-7]
? // [8-10]
( x >= 100000000ul // [9-10] [8]
? // [9-10]
( x >= 1000000000ul // [10] [9]
? 10
: 9
)
: 8
)
: // [6-7]
( x >= 1000000ul // [7] [6]
? 7
: 6
)
)
: // [1-5]
( x >= 100ul // [3-5] [1-2]
? // [3-5]
( x >= 1000ul // [4-5] [3]
? // [4-5]
( x >= 10000ul // [5] [4]
? 5
: 4
)
: 3
)
: // [1-2]
( x >= 10ul // [2] [1]
? 2
: 1
)
)
)
);
}
回答15:
If faster is more efficient, this is a improvement on andrei alexandrescu's improvement. His version was already faster than the naive way (dividing by 10 at every digit). The version below is constant time and faster at least on x86-64 and ARM for all sizes, but occupies twice as much binary code, so it is not as cache-friendly.
Benchmarks for this version vs alexandrescu's version on my PR on facebook folly.
Works on unsigned, not signed.
inline uint32_t digits10(uint64_t v) {
return 1
+ (std::uint32_t)(v>=10)
+ (std::uint32_t)(v>=100)
+ (std::uint32_t)(v>=1000)
+ (std::uint32_t)(v>=10000)
+ (std::uint32_t)(v>=100000)
+ (std::uint32_t)(v>=1000000)
+ (std::uint32_t)(v>=10000000)
+ (std::uint32_t)(v>=100000000)
+ (std::uint32_t)(v>=1000000000)
+ (std::uint32_t)(v>=10000000000ull)
+ (std::uint32_t)(v>=100000000000ull)
+ (std::uint32_t)(v>=1000000000000ull)
+ (std::uint32_t)(v>=10000000000000ull)
+ (std::uint32_t)(v>=100000000000000ull)
+ (std::uint32_t)(v>=1000000000000000ull)
+ (std::uint32_t)(v>=10000000000000000ull)
+ (std::uint32_t)(v>=100000000000000000ull)
+ (std::uint32_t)(v>=1000000000000000000ull)
+ (std::uint32_t)(v>=10000000000000000000ull);
}
回答16:
template <typename type>
class number_of_decimal_digits {
const powers_and_max<type> mPowersAndMax;
public:
number_of_decimal_digits(){
}
inline size_t ndigits( type i) const {
if(i<0){
i += (i == std::numeric_limits<type>::min());
i=-i;
}
const type* begin = &*mPowersAndMax.begin();
const type* end = begin+mPowersAndMax.size();
return 1 + std::lower_bound(begin,end,i) - begin;
}
inline size_t string_ndigits(const type& i) const {
return (i<0) + ndigits(i);
}
inline size_t operator[](const type& i) const {
return string_ndigits(i);
}
};
where in powers_and_max we have (10^n)-1 for all n such that
(10^n) < std::numeric_limits<type>::max()
and std::numeric_limits<type>::max() in an array:
template <typename type>
struct powers_and_max : protected std::vector<type>{
typedef std::vector<type> super;
using super::const_iterator;
using super::size;
type& operator[](size_t i)const{return super::operator[](i)};
const_iterator begin()const {return super::begin();}
const_iterator end()const {return super::end();}
powers_and_max() {
const int size = (int)(log10(double(std::numeric_limits<type>::max())));
int j = 0;
type i = 10;
for( ; j<size ;++j){
push_back(i-1);//9,99,999,9999 etc;
i*=10;
}
ASSERT(back()<std::numeric_limits<type>::max());
push_back(std::numeric_limits<type>::max());
}
};
here's a simple test:
number_of_decimal_digits<int> ndd;
ASSERT(ndd[0]==1);
ASSERT(ndd[9]==1);
ASSERT(ndd[10]==2);
ASSERT(ndd[-10]==3);
ASSERT(ndd[-1]==2);
ASSERT(ndd[-9]==2);
ASSERT(ndd[1000000000]==10);
ASSERT(ndd[0x7fffffff]==10);
ASSERT(ndd[-1000000000]==11);
ASSERT(ndd[0x80000000]==11);
Of course any other implementation of an ordered set might be used for powers_and_max and if there was knowledge that there would be clustering but no knowledge of where the cluster might be perhaps a self adjusting tree implementation might be best
回答17:
effective way
int num;
int count = 0;
while(num)
{
num /= 10;
++count;
}
#include <iostream>
int main()
{
int num;
std::cin >> num;
std::cout << "number of digits for " << num << ": ";
int count = 0;
while(num)
{
num /= 10;
++count;
}
std::cout << count << '\n';
return 0;
}
回答18:
Yet another code snippet, doing basically the same as Vitali's but employs binary search. Powers array is lazy initialized once per unsigned type instance. Signed type overload takes care of minus sign.
#include <limits>
#include <type_traits>
#include <array>
template <class T>
size_t NumberOfDecPositions ( T v, typename std::enable_if<std::is_unsigned<T>::value>::type* = 0 )
{
typedef std::array<T,std::numeric_limits<T>::digits10+1> array_type;
static array_type powers_of_10;
if ( powers_of_10.front() == 0 )
{
T n = 1;
for ( T& i: powers_of_10 )
{
i = n;
n *= 10;
}
}
size_t l = 0, r = powers_of_10.size(), p;
while ( l+1 < r )
{
p = (l+r)/2;
if ( powers_of_10[p] <= v )
l = p;
else
r = p;
}
return l + 1;
};
template <class T>
size_t NumberOfDecPositions ( T v, typename std::enable_if<std::is_signed<T>::value>::type* = 0 )
{
typedef typename std::make_unsigned<T>::type unsigned_type;
if ( v < 0 )
return NumberOfDecPositions ( static_cast<unsigned_type>(-v) ) + 1;
else
return NumberOfDecPositions ( static_cast<unsigned_type>(v) );
}
If anybody cares of further optimization, please note that the first element of powers array is never used, and the l appears with +1 2 times.
回答19:
in case the number of digits AND the value of each digit position is needed use this:
int64_t = number, digitValue, digits = 0; // or "int" for 32bit
while (number != 0) {
digitValue = number % 10;
digits ++;
number /= 10;
}
digit gives you the value at the number postition which is currently processed in the loop. for example for the number 1776 the digit value is:
6 in the 1st loop
7 in the 2nd loop
7 in the 3rd loop
1 in the 4th loop
回答20:
C++11 update of preferred solution:
#include <limits>
#include <type_traits>
template <typename T>
typename std::enable_if<std::numeric_limits<T>::is_integer, unsigned int>::type
numberDigits(T value) {
unsigned int digits = 0;
if (value < 0) digits = 1;
while (value) {
value /= 10;
++digits;
}
return digits;
}
prevents template instantiation with double, et. al.
回答21:
int numberOfDigits(double number){
if(number < 0){
number*=-1;
}
int i=0;
while(number > pow(10, i))
i++;
cout << "This number has " << i << " digits" << endl;
return i;
}
回答22:
// Meta-program to calculate number of digits in (unsigned) 'N'.
template <unsigned long long N, unsigned base=10>
struct numberlength
{ // http://stackoverflow.com/questions/1489830/
enum { value = ( 1<=N && N<base ? 1 : 1+numberlength<N/base, base>::value ) };
};
template <unsigned base>
struct numberlength<0, base>
{
enum { value = 1 };
};
{
assert( (1 == numberlength<0,10>::value) );
}
assert( (1 == numberlength<1,10>::value) );
assert( (1 == numberlength<5,10>::value) );
assert( (1 == numberlength<9,10>::value) );
assert( (4 == numberlength<1000,10>::value) );
assert( (4 == numberlength<5000,10>::value) );
assert( (4 == numberlength<9999,10>::value) );
回答23:
int numberOfDigits(int n){
if(n<=9){
return 1;
}
return 1 + numberOfDigits(n/10);
}
This is what i would do, if you want it for base 10.Its pretty fast and you prolly wont get a stack overflock buy counting integers
回答24:
int num,dig_quant = 0;
cout<<"\n\n\t\t--Count the digits in Number--\n\n";
cout<<"Enter Number: ";
cin>>num;
for(int i = 1; i<=num; i*=10){
if(num / i > 0){
dig_quant += 1;
}
}
cout<<"\n"<<number<<" include "<<dig_quant<<" digit"
cout<<"\n\nGoodbye...\n\n";
回答25:
for integer 'X' you want to know the number of digits , alright without using any loop , this solution act in one formula in one line only so this is the most optimal solution i have ever seen to this problem .
int x = 1000 ;
cout<<numberOfDigits = 1+floor(log10(x))<<endl ;
回答26:
This is my way to do that:
int digitcount(int n)
{
int count = 1;
int temp = n;
while (true)
{
temp /= 10;
if (temp != 0) ++count;
if (temp == 0) break;
}
return count;
}
回答27:
Here's a different approach:
digits = sprintf(numArr, "%d", num); // where numArr is a char array
if (num < 0)
digits--;
This may not be efficient, just something different than what others suggested.
来源：https://stackoverflow.com/questions/1489830/efficient-way-to-determine-number-of-digits-in-an-integer