cannot use power function from multiprecision boost library

[masterinex]

I have installed boost library version 1.65.0 . I cannot use the power function
"pow" from it. Whenever i write pow and use intellisense , it will only give me the pow function from math.h . I have included the following libraries:

#include <iostream>
#include <iomanip>
#include <sstream>
using namespace std;

#include <boost/multiprecision/cpp_dec_float.hpp>
#include <boost/multiprecision/debug_adaptor.hpp>
#include <boost/math/constants/constants.hpp>

#include <boost/multiprecision/detail/functions/pow.hpp>


using namespace boost::multiprecision;

int main() { cout << pow(112.2 ,3.4) << endl } ;

Are there some other libraries I can use to compute the power function of a large real number such as in the case of "sin(1.0)**2+cos(1.0)**2" so that it returns a value of 1.0 accurately ?

[Repeater]

Intellisense is not the compiler. How do you know that your program doesn't use the right pow function when it executes?

[Cubbi]

First, the expression pow(112.2, 3.4) will call the double, double overload of std::pow. If you want to call boost::multiprecision::pow instead, you need to have arguments of a type it understands, such as cpp_dec_float_50 or cpp_dec_float_100. Let's also bump the output precision:

#include <boost/multiprecision/cpp_dec_float.hpp>
#include <iostream>
using boost::multiprecision::cpp_dec_float_100;
int main() {
    cpp_dec_float_100 a = 112.2, b = 3.4;
    std::cout << std::setprecision(100) << pow(a, b) << '\n';
}

that prints

9332021.131197257591348458922719177023364501846775209425573316488942730002711710486442926388398457225

live demo: https://wandbox.org/permlink/tsy1Fv39kbpyJylr

Second, you want to work with high-precision math, but you're losing it right away:
the expression 112.2 is not 112 and two-tenths. It is a double with the value 112.2000000000000028421709430404007434844970703125. Likewise, "3.4" in a C++ program is not three and four-tenths, it is 3.399999999999999911182158029987476766109466552734375

So, fixing these issues,

#include <boost/multiprecision/cpp_dec_float.hpp>
#include <iostream>
using boost::multiprecision::cpp_dec_float_100;
int main() {
    auto a = cpp_dec_float_100(1122)/10;
    auto b = cpp_dec_float_100(34)/10;
    std::cout << std::setprecision(100) << pow(a, b) << '\n';
}

live: https://wandbox.org/permlink/86pIHR2prfIPHpLb

9332021.131197260700021602241138475068619624389028454081951744553393671746470878427399406755645506478

to compare, wolfram alpha for 112.2^3.4 prints
9.332021131197260700021602241138475068619624389028454081951744553393671746470878427399406755645506475578676563... × 10^6
which matches the second result to its second-to-last digit

Nearly all digits of the first result had nothing to do with 112.2^3.4

[masterinex]

ok , i just checked , it is actually using boost::multiprecision. however my final result is 1 not 1.0 , i think it looses the precision somewhere else, maybe at the addition of "sin(1.0)**2+cos(1.0)**2".