On 23/11/2007, Rich Shepard <[EMAIL PROTECTED]> wrote:
> Now I need to plot normal curves (a.k.a. Gaussian or bell curves,
> depending on the background of the speaker/writer). I see that SciPy has a
> class for the normal curve in its stats package, and that the curve shape is
> defined by the mean and standard deviation.
For parsimony, I think you're probably best off just using the
Gaussian equation:
def fwhm2k(fwhm):
'''converts fwhm value to k (see above)'''
return fwhm/(2 * n.sqrt( n.log( 2 ) ) )
def gauss1d(r, fwhm, c):
'''returns the 1d gaussian given by fwhm (full-width at half-max),
and c (centre) at positions given by r
'''
return exp( -(r-c)**2 / fwhm2k( fwhm )**2 )
(released to public domain)
> My need is to draw these curves based on the midpoint (== mean) and tail
> endpoints (which are not the same as the s.d.).
The midpoint here is c.
It's not clear what you mean by endpoints - if you mean you want to be
able to specify the y value at a given x delta-x away from c, then it
should be relatively simple to solve the equation to find the required
full-width at half-max to achieve these end-points. After a very quick
look (i.e. definitely needs verification), I think
k = sqrt( -(R-c)**2/log(Y) )
where Y is the desired value at distance R-c from the centre.
>Your thoughts are appreciated.
I hope that's what you're after.
Angus,
--
AJC McMorland, PhD Student
Physiology, University of Auckland
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Matplotlib-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
