Solving problem

Jbryant61 · ‎Jul 29, 2009

Hi. I have an equation that can take on different shapes (top hat or gaussian or anything inbetween). I have a desired size that I want the curve to take on. I want the size of the curve to reflect the 50% of the peak value. I'm having problems as when I change the size, the peak value also changes

any help, greatly appreciated.

Thanks
Jason

RichardJ · ‎Jul 29, 2009

For the half height point to occur at 16:

p(16,h)=p(0,h)*factor

Richard

Jbryant61 · ‎Jul 29, 2009

Thanks Richard/Jean.

Jbryant61 · ‎Jul 29, 2009

I've just realised this doesn't work if you normalise the original function by the total area (this was the original function Richard help with a few topics back)

Jason

RichardJ · ‎Jul 29, 2009

Could you post it. I don't want to hunt it down.

Richard

Jbryant61 · ‎Jul 29, 2009

Sure. Here it is.

I never asked but assumed the normalisation to area has something to do with the convolution and not having to mess about with constants?

thanks
Jason

RichardJ · ‎Jul 29, 2009

On 7/29/2009 12:11:32 PM, Jbryant61 wrote:
>Sure. Here it is.

See the attached

>I never asked but assumed the
>normalisation to area has
>something to do with the
>convolution and not having to
>mess about with constants?

If you have the convolution C=A*B, and the area of B is 1, then the area of C is equal to the area of A. So if B represents an instrumental broadening then that broadening redistributes signal, but does not change the total amount of signal. So the instrumental broadening function should always be defined in such a way that it's area is equal to 1.

Richard

Jbryant61 · ‎Jul 30, 2009

RichardJ · ‎Jul 30, 2009

On 7/30/2009 6:42:41 AM, Jbryant61 wrote:

>Was there supposed to be an
>attachment?

Sorry. Brain malfunction.

Richard

ptc-1368288 · ‎Jul 31, 2009

On 7/30/2009 6:42:41 AM, Jbryant61 wrote:
>Was there supposed to be an attachment?
>Thanks for the explanaition of unity area.
>
>Jason
>________________________

I have added this function to my collection of models, made it visible. BTW, 'b' is not needed as it can be any value once normalised, therefore the formula simplifies and 'b' is replaced by 1

jmG

Jbryant61 · ‎Jul 31, 2009

Its a very useful function - can swap between gaussain and top hat simply by the parameter "c". There was a problem though with defining "size" due to the shape change, but you and Richard have kindly sorted this out.

thanks to both of you!

Jason

ptc-1368288 · ‎Jul 29, 2009

Not sure all what you want, Jason.
At least, you can have Given/Find work.

Jean