I was given 10 data-points of time and the number of decays. I linearized them (by looking at $ln(A(t))$ where $A(t)$ is the number of decays) and wanted to compare linear regression, weightened linear regression and poisson regression.

For the Poisson regression I used the sekant method to find $a$ and $b$ for $y=a+bx$. I'd like to know the uncertainty of the a and b I found. For the linear regression we deduced a formula for the uncertainties by the gaussian error propagation, but this is not possible here as I have found $a$ and $b$ numerically here.

I'd like to know whether there's a better way than calculating the gaussian error for every step of iteration by taking partial derivatives for the used $a,b,x_i,y_i$ from the previous step.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.