EMASampler Class Reference

Exponential Moving Average (EMA) analog pin sampler. More...

#include <EMASampler.h>

Inheritance diagram for EMASampler:

Public Member Functions
	EMASampler (const uint8_t pin, const uint16_t periodMs=10, const float alpha=0.5)
	~EMASampler ()
virtual void	begin ()
virtual void	update ()
virtual float	average ()
virtual int16_t	last ()
Public Member Functions inherited from AbstractSampler
	AbstractSampler (const uint8_t pin, const uint16_t periodMs)
virtual	~AbstractSampler ()
	Destructor.
uint8_t	pin ()
	Accessor for the pin which is geting read from.

Protected Attributes
float	_alpha
	alpha value used in EMA calculation
uint32_t	_lastUpdated
	when last sample taken
float	_movingAverage
	most recently calculated mean value
int16_t	_lastSample
	keep the most recent sample
Protected Attributes inherited from AbstractSampler
const uint8_t	_pin
	pin to read data from
const uint16_t	_periodMs
	minimum ms

Detailed Description

Exponential Moving Average (EMA) analog pin sampler.

This sampler is used to calculate the exponential moving average for the analog values read from a pin. This method can be used to smooth a value over a large number of samples without having to use a lot of memory maintaining a buffer of recent values sampled from the pin, and should also be faster to calculate than the method used in the BufferedSampler approach. However, the user should be aware that the exponential moving average value is not the same as the mean value over a discrete set of samples (although it can often be used in situaltions where the mean may also be used).

Advantages:

• smoothing over long periods with low memory usage (low alpha values)
• effificient, constant time calculation of moving average
• responds faster to change than BufferedSampler

Disadvantages:

• not mean or median average (see link below)
• longer "tail" effect than BufferedSampler

See also: https://en.wikipedia.org/wiki/Moving_average#Exponential_moving_average

Definition at line 28 of file EMASampler.h.

Constructor & Destructor Documentation

EMASampler()

EMASampler::EMASampler	(	const uint8_t	pin,
		const uint16_t	periodMs = 10,
		const float	alpha = 0.5
	)

Constructor.

Parameters

pin	the analog pin to read samples from.
periodMs	the minimum gap between taking samples.
alpha	for determining number of samples to smooth over (between 0 and 1 - a higher alpha discounts older samples faster).

Definition at line 6 of file EMASampler.cpp.

~EMASampler()

EMASampler::~EMASampler ( )

inline

Destructor.

Definition at line 42 of file EMASampler.h.

Member Function Documentation

average()

virtual float EMASampler::average ( )

inlinevirtual

The mean value in the sample set.

Implements AbstractSampler.

Definition at line 63 of file EMASampler.h.

Here is the caller graph for this function:

begin()

void EMASampler::begin ( )

virtual

Initialization.

Should be called before use, and may also be called at any time to clear the moving average value. Note the moving average will be set to 0 when begin() in called.

Reimplemented from AbstractSampler.

Definition at line 15 of file EMASampler.cpp.

last()

virtual int16_t EMASampler::last ( )

inlinevirtual

Get the most recent sample.

Implements AbstractSampler.

Definition at line 67 of file EMASampler.h.

Here is the caller graph for this function:

update()

void EMASampler::update ( )

virtual

Time slice allocation.

This method should be called every time a sample is to be taken. If the object was constucted with periodMs > 0, this method will only take into account a new sample if at least periodMs milliseconds has elapsed since the last call to update().

Implements AbstractSampler.

Definition at line 24 of file EMASampler.cpp.

Here is the caller graph for this function:

---

The documentation for this class was generated from the following files:

• src/EMASampler.h
• src/EMASampler.cpp