On this page you will find all the exercise sheets for the computer exercises.
Some useful Root commands are collected in this Postscript file.
First steps with UNIX are summarised in an exercise sheet provided by B. Zittel
In this exercise, we will perform an unbinned Log-likelihood fit using the Minuit
minimisation package provided by ROOT
The exercise sheet can be found here: (PS) (PDF) . The supplemental exercise sheet is here: (PS) (PDF)
In this exercise, we will learn how to average several measurements of the same quantity.
In this example, a B hadron lifetime has been measured by 4 different experiments, each measuremnt is given with statistical and systematic errors. We will learn how to make an average of these numbers.
The exercise sheet can be found here: (PS) (PDF) .
You may use our template to get started (ASCII) (HTML) , as well as the Makefile (ASCII)
You can compare your result to mine
This project is concerned with reconstructing excided D-meson states. The data we will analyse has been recorded with the DELPHI detector, which was located at the LEP accelerator ring at CERN
An introduction can be found here (PS, 1.3MB) (PDF, 12MB)
The aim of the first exericise is to get aquainted with the ntuples, how to make cuts to enhance the signal and suppress the background and how to fit a Breit-Wigner function to an observed signal peak.
The exercise sheet can be found here: (PS) (PDF) .
You may use our template to get started (HTML) (ASCII)
This is another template which uses C++ classes. After downloading, unpack the archive with the command: tar -xzf DmesonProjekt.tar.gz
You can find a sample D0 mass peak with Breit-Wigner fit here.
Wolfgang prepared some code (ASCII) (HTML) you can use in your analysis/
The aim of this exercise is to combine several discriminating variables and form a likelihood variable
The exercise sheet can be found here: (PS) (PDF) .
You may use the following template as a basis for your analysis (ASCII) (HTML)
Those of you using Wolfgang's class can use the following modifications for this exercise: (ASCII) (HTML)
We will focus on learning how to use neural networks in an analysis in the final part of the DStar project.
I wrote a small neural network which is implemented as a C++ class. The network training routines are based on code originally written by O.Barring in Fortran. This is a very basic network, i.e. the fancy features described in literature or implemented in other networks are not available in this network. However, if trained properly, this network will have a performance suited well enough for the aim of this project. This network also has the advantage that the source code is rather short and you can understand how a network works without being distracted by too many features.
I have written a short introduction to neural networks in my diploma thesis which also contains some references for further reading.
The exercise sheet can be found here: (PS) (PDF) .
You will also need the following files:
Makefile (ASCII)
KaNet.cc (HTML) (ASCII) implementation of the class holding the neural network
KaNet.h (HTML) (ASCII) corresponding header
TrainNN.cc (HTML) (ASCII) example program for training the network
ApplyNN.cc (HTML) (ASCII) example program for using the trained network
ShowPlot.C (HTML) (ASCII) ROOT macro evaluating histograms filled during network training
