T O P I C R E V I E W |
AJOKi |
Posted - 09/10/2018 : 09:45:58 AM Hello,
I'm a new user of Origin, so far it's been very helpful to analyze my data however I'm not very good with math and I have some questions that I would like to ask about the Weibull function for data fitting. I have some experimental data that I want to fit to a mathematical model, so far the best model is the SWeibull2 function, as it is shown by the R2, X2 and fitting curve that appears after analysis with origin.
In the information that appears about the model, it shows the sigmoidal function: Y = A-(A-B)*exp -(kx)^d and mentions that has four parameters. However, I've reading about the Weibull model and came across different equations for this particular cumulative function, for example Y = Y0*(1-exp –(t-T/a)^b), or Y = 1-exp-(x/lambda)^k or Y = 1-(1-g)*exp-(kx/t)^b (they look similar but the exponential core function is different and I haven't been able to transform one into the other).
I was wondering if you could explain to me or link to me a couple of references where is clearly explained and in simplified terms if possible, from which four parameter cumulative Weibull function Origin derivates its SWeibull2 function to fit a set of data? Can you please explain step by stem parameterization to obtain Y = A-(A-B)*exp -(kx)^d ?
I understand that variables and parameter letters may change, and that the commonly referred "b" parameter in classical Weibull functions is the "d" parameter in Origin with the previous mentioned formula. If I were to write this formula on a scientific paper for the scientific community, Which form of notation is recommended?
I hope my questions are clear and I would appreciate very much any information you can give me.
Kind regards!
AJ-->thanks |
1 L A T E S T R E P L I E S (Newest First) |
Hideo Fujii |
Posted - 09/27/2018 : 4:45:46 PM Hi AJOKi,
I'm not a mathematician, but as far as I see that the standard form (F=1-exp(-(x/lambda)^k) in https://en.wikipedia.org/wiki/Weibull_distribution#Cumulative_distribution_function (k is the shape parameter, and λ is the scale parameter) is the case when A(top asymptote)=1, B(bottom at x=0)=0, and k=1/lambda in the Origin's formulation (y=A-(A-B)*exp(-(k*x)^d) ) in https://www.originlab.com/doc/Origin-Help/SWeibull2-FitFunc .
I hope math people can give further comment.
--Hideo Fujii OriginLab |
|
|