Struct AverageTime

template <int _L>

Exponential moving average for smoothing time-series data. More...

#include <structs.h>

Public Attributes

Type	Name
float	result = `0` Current average value.

Public Functions

Type	Name
void	Update (float _rate) Update average with new sample.

Detailed Description

AverageTime computes exponentially-weighted moving average (EWMA) to smooth noisy measurements like frame times or FPS. More recent values are weighted higher than older values.

Algorithm: * New average = Old average + (New sample - Old average) / Window size * Equivalent to EWMA with alpha = 1 / _L

Mathematical Properties: * Larger _L = slower response, smoother output * Smaller _L = faster response, more noise * Typical _L for FPS smoothing: 30-120 samples

Template parameters:

_L Window size (number of samples for smoothing)

Note:

Does not store sample history - constant memory usage.

Note:

Not thread-safe - use per-thread instances if needed.

Public Attributes Documentation

variable result

Current average value.

float AverageTime< _L >::result;

Public Functions Documentation

function Update

Update average with new sample.

inline void AverageTime::Update (
    float _rate
)

Incorporates new measurement into exponential moving average. Result converges to stable value after approximately _L samples.

Parameters:

_rate New sample value to incorporate

Note:

First _L samples have reduced accuracy (startup transient).

Note:

Can be called with any numeric type (implicitly converts to float).

Example:

AverageTime<60> fps_avg; // 60-sample window
while (running) {
    float frame_time = MeasureFrameTime();
    fps_avg.Update(1.0f / frame_time);
    DisplayFPS(fps_avg.result); // Smoothed FPS value
}

The documentation for this class was generated from the following file src/extra/structs.h