Date: prev next · Thread: first prev next last
2017 Archives by date, by thread · List index

Hi Brian:

        I've searched information in the web but usually it is more complex
than my knowledge can evaluate. For example, in Wikipedia found, but it so deep for me,
but show some graph with different values of Beta and Alpha that could
help me.

        I found some others like this, easier to understand the information:

        And found a program named "SciDavies" that could help me too.

        But the major problem to me is my knowledge in statistics. It is not to
deep, although I already found some books that could help me too.

        In general, when I made my question about it, I was trying to find a
way to make the process easier and fast to me.

        Thank you very much because you try to help me, and clear me that
LibreOffice can't help with its formulas to solve my problem directly.



Jorge Rodríguez

El mar, 09-05-2017 a las 16:25 +0100, Brian Barker escribió:
At 08:08 08/05/2017 -0600, Jorge Rodríguez wrote:
Hi Brian, well, thank you for your answer

My point, of course, was that it was sufficient - 
indeed, better - to explain that you had a table 
with consecutive integers in one column and real 
values in the other, rather than listing 
eighty-two values and asking readers to make that observation themselves.

At 14:10 07/05/2017 -0600, Jorge Rodríguez wrote:
If I have a number of data about some topic, how 
can I get the type of distribution using the 
LibreOffice's formula ? How can I know if it is 
normal, gamma, etc. ? There is a way ?

The simple answer is that your data will 
certainly match none of those distributions - at least, not precisely.

Think of a straight line: this can be defined 
exactly by two points. If you have three points, 
they will not - except by an impossible 
coincidence - lie on a straight line. But if you 
believe that your data should lie on a straight 
line, there are techniques that determine the 
straight line which probably misses all of your 
points but is in some sense the "best fit" to them.

Similarly, your analytic distributions will be 
definable by a small number of points - certainly 
a lot less than your forty-one. So the best you 
can do is to attempt to fit various distributions 
to your data and also to discover how good the 
fit is in each case. That might help you to see 
which distribution the data fits best. But a 
better logic would be to decide from the nature 
of your data - what it represents - which 
distribution is likely, and then to fit the data 
to such a distribution (if that is necessary).

In any case, Calc does not appear to have any 
functionality that would help you here, but you 
will certainly be able to build formulae to 
achieve what you want in Calc once you have 
unearthed the relevant mathematics. That will surely be out there on the web.

I trust this helps.

Brian Barker

To unsubscribe e-mail to:
Posting guidelines + more:
List archive:
All messages sent to this list will be publicly archived and cannot be deleted


Privacy Policy | Impressum (Legal Info) | Copyright information: Unless otherwise specified, all text and images on this website are licensed under the Creative Commons Attribution-Share Alike 3.0 License. This does not include the source code of LibreOffice, which is licensed under the Mozilla Public License (MPLv2). "LibreOffice" and "The Document Foundation" are registered trademarks of their corresponding registered owners or are in actual use as trademarks in one or more countries. Their respective logos and icons are also subject to international copyright laws. Use thereof is explained in our trademark policy.