Particle physics lab: techniques

[audreyh]

Hi all,

I'm measuring multiple coulomb scattering by using 4 chambers filled with He gas, and lead plates between the top 2 chambers and bottom two. The chambers have a potential difference across the plates so when the muons ionize the He gas, a spark is produced. A camera marks the pixels of the sparks.

Homework Statement

The computer program written in root plots number of events (y) vs scattering angle (x). The fit to the data points is a double gaussian.

$f(x)= e^{- \frac {x^2} {2 \sigma_1^2}} + e^{- \frac{x^2} {2 \sigma_1^2}}$

We'd like to see, for more lead plates a larger scattering angle distribution (i.e. Sigma to increase). The problem is, the experiment with zero lead plates produced non-zero scattering angles (expected), so I'd like to subtract the straight through data (zero lead plates) from the data that has a non-zero number of lead plates. This is what I'm not sure how to do.

Homework Equations

The Attempt at a Solution

From my limited understanding of probability: if X is a continuous random variable that is normally (gaussian) distributed with parameters mu and sigma^2, then
$Y= \alpha X+ \beta$
with parameters
$\alpha \mu + \beta$
and
$\alpha^2 \sigma^2$
So can I somehow linearly convert the non-zero lead plate distributions to one that has the zero-plate scatter distribution subtracted?

 

[Jhero]

Hi there,

It sounds like you are trying to subtract the data from your experiment with zero lead plates from the data with non-zero lead plates in order to isolate the effect of the lead plates on the scattering angle distribution. This is a common technique used in data analysis and can be done by subtracting the two histograms (or curves in this case) from each other.

To do this, you will need to make sure that the two histograms have the same binning in order for the subtraction to make sense. Once you have done this, you can simply subtract the y-values for each bin in the histogram with zero lead plates from the y-values for each bin in the histogram with non-zero lead plates. This will give you a new histogram that represents the difference between the two distributions.

Alternatively, you can also fit a curve to the histogram with zero lead plates and then subtract this curve from the histogram with non-zero lead plates. This will give you a more precise subtraction, but may require more advanced data analysis techniques.

I hope this helps and good luck with your experiment!