How to decimate temporal data in Excel? That is to reduce the number of
samples to the minimum significative number.

I am doing so by hand averaging groups of x-samples and checking visually if
the data to be averaged is significative to the meaning or simple noise.

Doing so over a large number of data is more than cumbersome, it is a
non-sense. I can code it but I need to teach to students how to decimate and
they don't necessarily know how to code.

Thank you very much for your help

Afraid I don't know too much about decimation/reducing noise, you'll need to
explain the math/logic needed for it.
OR
Try looking up some of the stat math functions in Excel's help, such STDEV.
--
Best Regards,

Luke M
*Remember to click "yes" if this post helped you!*


"Guady" wrote:

How to decimate temporal data in Excel? That is to reduce the number of
samples to the minimum significative number.

I am doing so by hand averaging groups of x-samples and checking visually if
the data to be averaged is significative to the meaning or simple noise.

Doing so over a large number of data is more than cumbersome, it is a
non-sense. I can code it but I need to teach to students how to decimate and
they don't necessarily know how to code.

Thank you very much for your help

On Dec 15, 8:58 am, Guady wrote:
I am doing so by hand averaging groups of x-samples and
checking visually if the data to be averaged is significative
to the meaning or simple noise.

If you do a google search for "outliers", you might find some
procedures for recognizing and removing outliers that will appeal to
you.

However, most knowledgable staticians deprecate automated methods of
removing outliers from the data.

Instead, it is appropriate to use automated methods to __identify__
outliers, but it requires manual interpretation to determine if the
outlier should be removed (generally, only if the outlier represents
erroneous data, e.g. a probable data entry error). Non-erroneous
outliers might be necessary to properly define the distribution, even
if they lie outside +/- 3 or 4 sd or +/- 3 or 4 IQR (interquartile
range), to mention two of the many methods for recognizing outliers.

Obviously, that is indeed a manual process. So you might be doing the
right thing, and that is what you should teach your students.

That said, perhaps TRIMMEAN is the Excel function you are looking.
IMHO, that is the poorest way to define and deal with
"outliers" (not!). But the fact is, some people do indeed (mis)use
TRIMMEAN for that purpose.

(IMHO, TRIMMEAN is useful only for methods of measuring performance
that are defined in the manner. For example, a similar method is used
for scoring in some Olympic events, as I recall.)


----- original posting -----

On Dec 15, 8:58*am, Guady wrote:
How to decimate temporal data in Excel? That is to reduce the number of
samples to the minimum significative number.

I am doing so by hand averaging groups of x-samples and checking visually if
the data to be averaged is significative to the meaning or simple noise.

Doing so over a large number of data is more than cumbersome, it is a
non-sense. I can code it but I need to teach to students how to decimate and
they don't necessarily know how to code.

Thank you very much for your help