Awesome

ScaledValue

A lightweight, compile-time, single header-file library class for representing (eg double-) values on different scales built on c++11 with no dependencies.

Latest Release - v1.0.0

Initial release.

Tested on

• gcc-5.3.0

Compiles Under

• icc-16 + 17
• gcc-7 .. gcc-4.7
• ARM gcc 5.4
• ARM64 gcc 5.4
• MIPS gcc 5.4
• clang 4.0 .. clang 3.4

Contents

• ScaledValue
• Latest Release - v1.1.0
  • Tested on
  • Compiles Under
• Contents
• Description
• Limitations
• Getting Started Guide

Description

The library allows to represent a (e.g. double) value on a scale. The following example illustrates the idea:

#include <ScaledValue.h>

int main() {

    using namespace sv;
    
    //define values on different scales
    auto seconds_a = ScaledValue<std::nano>{500}; // represents 5 nanoseconds
    auto seconds_b = ScaledValue<std::micro>{9};  // represents 9 microseconds

    //combination
    auto seconds_c = seconds_a + seconds_b;
    auto seconds_d = seconds_a * seconds_b;
    bool greater = seconds_a > seconds_b; //greater==false

    //math operations
    auto seconds_e = math::pow(seconds_b,0.5); 

    //conversions
    auto seconds_f = ScaledValue<std::nano>{seconds_b};

    //printing
    std::cout<< seconds_b <<std::endl; // prints "9*micro"
    std::cout<< seconds_c <<std::endl; // prints "9500*nano"
    std::cout<< seconds_d <<std::endl; // prints "0.0045*nano"
    std::cout<< seconds_e <<std::endl; // prints "2645.75*micro" // see *
    std::cout<< seconds_f <<std::endl; // prints "9000*nano"
    
    return 0;
}

Working on a specific scale incurs no runtime overhead (on -O1). The library is small (~200 sloc), tested, easy to understand and easy to extend to specific user needs. It can be used together with the units library.

*Note that since the true value of seconds_b is actually 9e-6 (seconds), the root is taken from this value rather than from 9 yielding 0.00264575 (seconds) and not 3.

Limitations

Scales are compile time properties so they cannot be used in a dynamic fashion*. Printing of scaled values is rather simple but can be adjusted if needed. Not all math functions from <cmath> are defined, but they are also simple to add.

• Dynamic behaviour can be introduced by using std::variant (c++17) with this class.

Getting Started Guide

The library consists of a single file (ScaledValue.h). To incorporate the library into your project, simply copy the header into a location in your include path and include it as seen in the Description.