Proposal: AverageValueInstrumentSample --> MeanValueInstrumentSample

[Ryan Shaw <ry...@stanfordalumni.org>]

Hi,

I would like to propose that AverageValueInstrumentSample (in the
instrument.manager package) be renamed to MeanValueInstrumentSample
before a release is done. This would also involve changing the 
InstrumentSampleFactory to accept the type value "mean" rather 
than "average".

Why? Because the AverageValueInstrumentSample is actually calculating
a mean. "Average" is a vague term which includes mean, but also includes
other measures such as mode and median [1]. While it is true that people
ususally mean "mean" when they say average, I think we should strive
to be as precise as possible. This will also allow the creation of other
ValueInstrumentSamples which measure other averages such as median and
mode, without naming conflicts.

Of course even "MeanValueInstrumentSample" may not be precise enough,
as there are differents types of means [2]. I haven't looked yet at what
type of mean it is calculating...we may want to call it 
ArithmeticMeanValueInstrumentSample if it is simply calculating the 
arithmetic mean.

Ryan

[1] http://www.shodor.org/interactivate/dictionary/a.html#average
[2] http://mathworld.wolfram.com/Mean.html

--
To unsubscribe, e-mail:   <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>

Re: Proposal: AverageValueInstrumentSample --> MeanValueInstrumentSample

[Berin Loritsch <bl...@apache.org>]

Leif Mortenson wrote:
> Ryan Shaw wrote:
>>
> :-)  I'm just an EE geek.  It is caluclating the average by taking the 
> total of all values in a sample
> period  divided by the number of values in that period.


Yep.  That's a Mean.  The Average would be for all samples in the run.

> 
> The dynamic sample stuff is still in flux as I am trying to find time to 
> complete it so I am open to
> opinions here.


Which is good.

Now, how about an RMS sampler? :)

Being an Audio Geek, Root Mean Square average values show a closer
representation to perceived loudness.

             /------------------
            /  --
RMS =     /   \   x^2
          /    /
        \/     --

(hows that for ascii math representation?)

Actually it isn't quite right:

RMSn1..3 = sqrt( ( Xn1^2 + Xn2^2 + Xn3^2) / n );

Where Xn is the sample number, n is the number of samples



-- 

"They that give up essential liberty to obtain a little temporary safety
  deserve neither liberty nor safety."
                 - Benjamin Franklin


--
To unsubscribe, e-mail:   <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>

Re: Proposal: AverageValueInstrumentSample --> MeanValueInstrumentSample

[Leif Mortenson <le...@tanukisoftware.com>]

Ryan Shaw wrote:

>Hi,
>
>I would like to propose that AverageValueInstrumentSample (in the
>instrument.manager package) be renamed to MeanValueInstrumentSample
>before a release is done. This would also involve changing the 
>InstrumentSampleFactory to accept the type value "mean" rather 
>than "average".
>
>Why? Because the AverageValueInstrumentSample is actually calculating
>a mean. "Average" is a vague term which includes mean, but also includes
>other measures such as mode and median [1]. While it is true that people
>ususally mean "mean" when they say average, I think we should strive
>to be as precise as possible. This will also allow the creation of other
>ValueInstrumentSamples which measure other averages such as median and
>mode, without naming conflicts.
>
>Of course even "MeanValueInstrumentSample" may not be precise enough,
>as there are differents types of means [2]. I haven't looked yet at what
>type of mean it is calculating...we may want to call it 
>ArithmeticMeanValueInstrumentSample if it is simply calculating the 
>arithmetic mean.
>
:-)  I'm just an EE geek.  It is caluclating the average by taking the 
total of all values in a sample
period  divided by the number of values in that period.

The dynamic sample stuff is still in flux as I am trying to find time to 
complete it so I am open to
opinions here.

Cheers,
Leif



--
To unsubscribe, e-mail:   <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>