# I Likelihood of some points fitting a derived function...

Tags:
1. Sep 9, 2016

### Silviu

Hello! I have some concentrical spheres with many points inside. And I need to plot the total mass of points in each shell (so between 2 spheres) versus the radius of that shell (defined as (r1+r2)/2, where r1 and r2 are the radius of the 2 spheres forming the shell). I have 10 shells so my plot has 10 points and I want to fit them with a certain function and I need to calculate the likelihood associated with this fitting function. But for this, I need a sigma and I don't know what exactly would that be. So i know the mass, radius and number of points in each shell. I thought to take the error something like sqrt(number of particles) but I am not sure. What should I do to calculate the likelihood?

Thank you!

2. Sep 9, 2016

### Stephen Tashi

What is the probability model for the situation ?

Unless you can state a mathematical model, you'll only get vague and general advice.

"Liklihood" is technical term in statistics, and I don't know whether you intend to use it in a technical sense. It isn't clear what random variable you are talking about when you speak of "sigma". If the data in each shell follows a different probability distribution, the data in different shells might be realization of different random variables, each with its own standard deviation.

3. Sep 10, 2016

### chiro

Hey Silviu.

You should provide the model you want to use to fit the data and then provide the assumptions for the original data you are using.

After that, it's a matter of going through the steps [either using things like Central Limit Theorem if you have lots of data or using specific techniques if this is isn't the case].