Re: How to resample continuosly time data to be represented real time ?

Here is a possible idea:

The final result will be an average for each line/point though:

First simply collect 500 measurements.

Then for each next measurement.

Two measurements need to be average... so two become one.

At first the most left becomes the average... of measurement 0 and 
measurement 1.

Measurements 2 to 500 remain intact.

Now keep track of an average index which specifies which element of the
500 
is to be average next.

So the average index goes to 1.

Now a new measurement comes in...

Value at average index1 is averaged with value at index 1.

The rest is ****fted down.

So what will happen is something like this, let's say it's limited to 10 
measurements:

0123456789  // 10 measurements

0
1234567890 // 11 measurements
N

02
1345678901 // 12 measurements
NN

024
1356789012 // 13 measurements
NNN

0246
1357890123 // 14 measurements etc.
NNNN

           // N = the new value of the previous 2 averaged.

// Example:
0+1/2
2+3/2
4+5/2
6+7/2

Then when it's full... when the average index reaches the last it goes
back 
to zero.

So then it goes like so:

0246802468
1357913579
----------
NNNNNNNNNN   // 20 measurements


0
1234567890 <- N  // 21 measurement
         |
         21

Now it starts re-average the stored value.

There are two ways to do this:

(0 + 1) / 2 = Average... this is a low pass filter I think they call it.

Another better way would be to keep track of how many lines were averaged.

(=How many times did the average index go back)

So then the formula becomes:

( ( OldN * (AverageLines-1) ) / AverageLines ) + ( NewMeasurement / 
AverageLines ) = NewN

You get the idea ?

This way all measurements end up in the graph averaged and none are thrown

away ! ;)

Bye,
  Skybuck.