# Exercise: `RandomSensor_nopoly` (Non-Polymorphic)¶

## Requirement¶

Implement a sensor class `RandomSensor_nopoly` which does not access any real hardware, but rather returns a random temperature within a given range.

```#include <gtest/gtest.h>
#include <sensor-random-nopoly.h>

TEST(sensor_random_suite, basic)
{
RandomSensor_nopoly rs(36.4, 42.3);                // <--- give measurements between lo and hi

double t = rs.get_temperature();
ASSERT_GE(t, 36.4);                                // <--- greater equal to lo
ASSERT_LE(t, 42.3);                                // <--- less equal to hi

ASSERT_FLOAT_EQ(rs.low(), 36.4);                   // <--- RandomSensor specific interface
ASSERT_FLOAT_EQ(rs.high(), 42.3);                  // <--- RandomSensor specific interface
}
```

### Implementation Hint: Randomicity In C++¶

• The following program demonstrates how to generate uniformly distributed random `double` values in a specified range.

```#include <random>
#include <iostream>

int main()
{
// setup how we want our random numbers to be created (think:
// object initialization)
double lo = 36.1;
double hi = 42.7;

// object state. think: what does the object need during runtime?
std::random_device rd;                                       // <--- /dev/random, used to obtain a seed
std::default_random_engine engine(rd());                     // <--- init (and seed) RNG
std::uniform_real_distribution<double> distribution(lo, hi); // <--- tailor desired distribution

// draw 10 random numbers (think: measure temperature)
for (int i=0; i<10; i++) {
double number = distribution(engine);                    // <--- use uniform_real_distribution to
//      draw a number from engine
//      and bend it to desired range

std::cout << number << std::endl;
}

return 0;
}
```

### Implementation Hint: Initializer Lists¶

See Constructors: Member Initialization for how to initialize members in an initializer list.