Final+Project

For my final project, I chose to use the characteristics I have learned about cosmic rays through various mechanisms--scatter plots, histograms, etc-- to isolate cosmic rays from a 100k event set of dimuon data.

The first of these characteristics about which I learned is their mass. We saw time and time again in scatterplots and histograms we noticed breaks and changes in the slopes at 40 GeV. After considering the characteristics that we know about cosmic rays as well as the way we would expect the data to appear without any cosmic rays, we were able to deduce that this was the point at which cosmic rays began tampering with the data. Below is a photograph of an eta2 vs mass plot. Knowing that cosmic rays should enter from space into the detector in a straight line, they should appear on the Eta 1 graph and Eta 2 graphs at Eta=0. Below is a graph of eta2 vs mass. I used one of the ManyEyes features to highlight some of the particles along this line. In the graph above, I switched from Eta2 to Eta1. Before 40 GeV, the particles that I had selected scattered completely. This is how we would expect the intended particles in the collision to behave. After 40 GeV, many of the particles continued along a straight line. These, we can assume, are cosmic rays because they continued a straight trajectory.

The histogram below shows the number of events that occur at various mass values. From 0-30 GeV, the number of particles that possess each mass gets progressively smaller until it dwindles to 0. After having a value of 0 from 30-40 GeV, there is a spike in the number of particles right at 40.

These are just a few of the graphs that led me to believe 40 GeV is the "magic number" at which cosmic rays begin to appear. The first cut I made on the excel spreadsheet of data was one based on the data sorted by mass and cut after 1055 events.

The next criterion I used to cut the data is its energy to mass ratio. We know that E^2=m^2+p^2. Because of the way the detector reads a single cosmic rays as 2 separate particles, the momentum of the "two halves" cancels the other out, so momentum (p) is equal to zero. This leave E^2=m^2, or E=m. (E1+E2)/M, then, should be close to equaling 1. I created a new column for these values, sorted by this column, and cut accordingly. Here is a portion of the cut:

A third characteristic is the phi1 and phi2 values. Because of the way the phi coordinate system functions, we would expect the straight trajectory of a cosmic ray to provide values of positive and negative 1/2 pi on phi 1 and phi 2, respectively. The sum of phi 1 and phi 2, then, should be zero if the particle travels through the detector in a straight line. This graph shows that at 40 GeV, the sum of phi 1 and phi 2 begins to produce a value close zero. 

I created another new column of phi1+phi2 values, sorted by this, and cut to include all the values between -0.50 and 0.50



The final criteria that I used in my cuts of the data is that of the charge, represented by Q. The detector assumes that the cosmic ray, read as two separate particles, possesses opposite charges (in other words, that each of the perceived halves has a different charge). Due to this, the value of Q1*Q2 should be an overall negative value. (negative times positive is negative). I created a new column of the Q1*Q2 values, and then used my own ManyEyes account with my uploaded data to create a scatterplot of this vs. event number. This provided an additional several particles to remove from the data set. 

After all of the cuts were made, I was left with events that we can reasonably assume are cosmic rays because they match all of the criteria. I was able to successfully isolate cosmic rays from the 100,000 event set of data based on criteria that we learned and deduced over the semester.