uint get_factorial(uint nValue)
{
	if( !nValue ) return 1;
	for(uint i=nValue; i>1; i--,nValue*=i);
	return nValue;
}

Any other questions, let me know.

Origin's built-in function list includes a function called gammaln(x). This function represents the natural log of gamma(x). The reason why Origin provides only the natural log of gamma has to do with computational limitations. In other words, in computing gamma functions, one can quickly run off into very large numbers. To avoid this, Origin internally computes ln(gamma(x)) rather than gamma(x). (Note that this is the same reason that we don't provide a built-in factorial function in the first place...factorials also blow up very easily.)

So, using the gammaln() function, your user-defined function would be as follows:

y=(P1^x/exp(gammaln(x+1)))*exp(-P1)

The Form can now be set to equation. (Note that you don't need the Y-Script Form anymore.)

Any questions, please let me know. I'd also be happy to hear when you get it working.

However, you can write a script to calculate the factorial. The script can be found in the Knowledge database, http://www.originlab.com/www/tech/resultstech.asp?ItemID=241&VersionID=All&TitleID=Script+to+calculate+the+factorials+< !--+LabTalk+math+dataset+-->&CatID=Worksheet+Windows+/+Datas ets&TypeID=Script">Script to calculate the factorials.

With the ability to write a script you can then define a user-defined fitting function of Form, Y-Script. To do this, please follow these steps:

1. Open the NLSF, Analysis:Nonlinear Curve Fit. (You should be running in the Advanced Mode. You know that you are in the Advanced Mode if you see the Basic Mode button.)
2. Select Function:New.
3. In the Form drop-down list, select Y-Script.
4. Enter in your Definition as the following:

num=x;
res=num;
if(num > 1)
for(res=num;num>1;num-=1,res=res*num);
y=(p1^x/res)*exp(-p1);

5. Continue as you normally would when fitting.

Laurie Shea
Technical Support