What Does True Mean in Particle Simulation Data?

[harten]

Hi there!
I'm new here and have a pretty basic and maybe stupid question:
What exactly does one mean when talking about "truth information"?
I'm working with a simulation of a particle detector and I've been trying to make sense out of all the data that is stored in the ouput files. Words like "true hit position", "true signal" and "truth information" are coming up. But I don't really get what it means.

 

[Orodruin]

Reference please. When you make statements such as "what does it mean to say x", it is taken completely out of context and we will generally have no idea about how to interpret it.

 

[harten]

I thought the context I provided would be enough... Here's some more:
I have a simulation of a particle detector (as I said) and there is a peace of software that writes the simulated events in a (root) file. There things like "true hit position" etc. are saved. I was wondering why one uses the word "true" and what it stands for in this context. I googled around but couldn't find anything...
Thanks!

EDIT: Sorry, the simulation is done with geant4

 

[Vanadium 50]

Reference please.

Show us where you see it. Don't just repeat the claim. Let us read the original so we understand the context.

 

[Orodruin]

I thought the context I provided would be enough... Here's some more

You cannot just go around assuming that people know what you are talking about. You need to provide the full reference to where you have read this, not just quote selected parts of it. This is what "providing a reference" means.

 

[Hepth]

He said Geant4. I'll take a look and see if I can find something for you.

 

[mfb]

In general, "true" is what the simulation used as variable (like the position where the particle went through the detector layer in the simulation - this will be different from the reconstructed position).
For specific cases, more context is necessary.

 

[Hepth]

I think it has to do with the MC particle generation. Sort of like learning, your MC (Monte Carlo) simulator for the event will have a "True X" that gets fed into your simulated detector. The simulated detector then has a "reported X" that may or may not match the "True X". So the MC knows the true hit position, but that could be different from your detector's reported hit position.

EDIT : as mfb just pointed out

 

[fzero]

It seems that the term "truth information" refers to the generated events, consisting of an initial state, intermediate state, and final state, of a Monte Carlo simulation. For real detector physics, we might know the initial state (the beams) with some precision. The final state will be known imperfectly, since we will have uncertainties in the particle id, energy and momentum measurements, as well as particles like neutrinos that pass through the detector without measurement. Furthermore, the intermediate state is not even a physical observable and there will be many intermediate states contributing quantum mechanically to a given event. Nevertheless, in a physics experiment we do want to use the physical observables of the final state to reconstruct the possible intermediate states contributing to the event. In the detector simulation we can use the truth information to teach us about this reconstruction process.

Some discussion about the truth information can be found in Part III of http://arxiv.org/abs/1101.2599 and sect. 14.2 of http://arxiv.org/abs/1003.1643, but you might try to find other MC technique sources to explain the more basic reconstruction techniques that I am not familiar with.

 

[Vanadium 50]

Ah, the old PF game of "guess what an unclear post means".

He's asking us what a phrase means in a particular context. Doesn't it make sense to find out what that context is before starting the guessing?

 

[Hornbein]

Hi there!
I'm new here and have a pretty basic and maybe stupid question:
What exactly does one mean when talking about "truth information"?
I'm working with a simulation of a particle detector and I've been trying to make sense out of all the data that is stored in the ouput files. Words like "true hit position", "true signal" and "truth information" are coming up. But I don't really get what it means.

If it is a simulator, I'd guess it comes up with a random hit position, the "true hit position", then adds in some error. Same with the signal. But "truth information" could be anything.

 

[harten]

Thank you mfb, Hepth and fzero!
I'm sorry, some of you think, that the context was not enough, but there was no more significant context. The information these guys provided were exactly what I was looking for. Thank you again!

 

[Orodruin]

but there was no more significant context

In order to be in a scientific environment, you will have to learn how to properly reference things. You cannot simply say that "I read this", you need to specify where you read it and provide a reference or link so that people can go to the original source to find the context. It does not really matter what you consider to be significant context. The very fact that you are asking about it means that you might be missing context when quoting.

 

[Hepth]

But his second post he said was using geant4. He's in a particle forum, talking about detector and simulation output file properties. I'm not sure why you all jumped on him for "context", seemed mfb, fzero and I picked up on it pretty fast?

He basically asked "When doing a particle event simulation, in a geant4 output file there are properties listed as "true X". Does anyone know what "true" means in this case?"