Re: Calculating mean and variance for normal distribution

[Ben Abbott]

On May 24, 2008, at 12:02 PM, DimitryASuplatov wrote:

> Hello,
> I have a common task - to analyze some data, presumptively normaly
> distributed, and calculate its mean, variance and make a plot. I`ve
> checked the book-manual and the mailing list, but all functions that
> have something to do with normal distribution require mean and  
> variance
> as an input parameters. This is very strange. I feel like missing
> something...
>
> Could you just give me a tip. Lets say I have my data in vector x. How
> to get mean and varience assuming normal distribution? And if this  
> would
> not cause many trouble could you give me a link to where I can read  
> this
> because I might be reading in the wrong place.
>
> Thanks a lot for your help!!!
>
> SDA.

I'm a bit confused by your question. The definition of mean and  
variance are each independent of distribution, are they not?

In any event, at Octave's prompt type

> help mean
> help std

Regarding a plot, I'm not sure what you hope to do. Are you looking to  
plot a histogram with fitted normal distribution superimposed over it?

If so, to calculate the pdf of a uniform distribution

> help normpdf

Regarding a reference, the link below might be what you're looking for.

        
http://www.gnu.org/software/octave/doc/interpreter/Statistics.html#Statistics

Octave's statistics function are generally compatible with the "other"  
programs ;-)

Thus, the link below might also be of use to you.

        http://www.aquaphoenix.com/lecture/matlab10/page2.html

Ben