|
|
|
PHYS 4807/5002: Computational Physics - Statistical Data Analysis
Fall 2014
Professor Alain Bellerive
Carleton University
alainb [at] physics.carleton.ca
Description:
The course Physics 4807 provides a thorough introduction
to the statistical methods used in experimental physics. Consequently
it is a course on applied statistics for data analysis. It will start
with a brief overview of the ROOT data analysis framework (from CERN)
followed with an introduction to the UNIX operating system. The
main software and graphics package will be ROOT (C++ based). Students
can also used older packages such as PAW (from CERN) or GRPACK (by S. Brandt), which are Fortran based.
First the concept of histograms and then
probability will be stated
for computational methods used in analysis of experimental data. Then
the course will cover: Simple linear fit - Monte Carlo methods for simulation of random
processes - Important distributions and theorems - Statistical methods
for parameter estimation and hypothesis tests - Numerical methods
for solving problems in linear algebra, integration, and minimization
- Probability, confidence intervals, stochastic calculations - Maximum
Likelihood and Least Squares methods - Multivariate data classification
(e.g. likelihood functions). Examples will be
taken primarily from applied physics.
An overview of programming hints, compilation tricks, shell languages, and object oriented
framework in C++ for the data analysis package ROOT will be provided at the beginning of the class.
Students can use other languages such as Fortran, Pascal or Java, but C++ is strongly recommended. Students
are discourage the usage of Excel or Mathematica for this course (those tools have
limitations for advanced analysis and manipulation of data).
Also offered at the graduate level as Physics 5002
with additional requirements. Prerequisite: Permission of the department
and an ability to program in Fortran or C.
Concepts:
The course provides a thorough introduction to the
statistical methods used in experimental physics. In Computational
Physics, students apply simple computer programming skills to the analysis
of data from physical systems. The focus applied statistics and data handling.
Fundamental concepts in probability
theory are developed. Statistical methods, including hypothesis tests,
parameter estimation using the maximum likelihood and least squares
method, and confidence interval calculations, are applied. The simulation
of physical systems using Monte Carlo techniques is explored. The
course teaches students how to quantify uncertainties in complex situations,
which is important in physics, engineering and in other areas of science.
For example in relating models to actual data and systems to risk analysis.
Why computational physics?
Advanced students in science need to understand
the basic data handling methods so that they can make effective use
of the powerful computer programs that are now widely distributed
for analyzing large experimental data sets. There is probably no need
to convince any physicist that computers and fast networking are changing
the way we address data analysis. Many problems encountered in science
involve complex systems that can be solved with numerical techniques.
Computational physics is the discipline which encompass the rules
for the development of efficient algorithms to provide qualitative
descriptions of a physical system with the proper error analysis and
the most complete data reduction scheme for a set of quantitative
measurements.
Some Advice on How to Succeed in this Course
This course will be different from many of your upper
division physics courses. You may find it hard to keep track of all
the new terminology and ideas. Here is some advice on how to deal
with it.
-
Come prepared at each lecture! Print
the lecture notes ahead of time. Student who does not show-up
in class will not be allowed during regular office hours.
-
Keep up with the reading and do the homework
on time. Take careful notes when you read the textbook and
programmed the examples. Come to office hours to get mysterious
concepts clarified.
-
Remember information such as estimators
for the distributions of a sample, principles for the least squares
and maximum likelihood fitting, arguments for testing a fit, procedures
for parameters estimation, and rules for the computation of confidence
intervals & test hypotheses.
-
You will need to learn and understand the
concept of probability and statistics. In order to provide
an accurate evaluation of the data, you must be able to put the
information into perspective. Without the background information
in your head, you will have a hard time understanding this context.
Creating your own mental database also helps you to develop physical
intuition in a subject that is unfamiliar.
-
Remember the main results of homework problems.
Many of the problems will address important issues. Use this web
page as a guide.
-
Data Analysis deals with data reduction
and proper error analysis. Do not focus on the subtle aspects
of the language used for a given routine, but instead focus on
the algorithm and the methodology.
-
Graduate level: Graduate students registered
to 5002 will have extra problems in some assignments and will be
asked to cover more material.
- Academic Policies: Refer
to the Department of Physics
Lectures
Midterm and Final (example from a previous year)
- Midterm 2006 (pdf)
- Final 2006 (pdf)
- Formula Sheet (pdf)
Virtual Machine on Windows
Install ROOT on a MAC
Code and Links
Homework
- Homework: One every week (or every two weeks), due IN CLASS 7 days
(or 14 days) after it was assigned (assignment policy).
- Assignment #1 (due September 17, 2014)
- Assignment #2 - Data (due September 25, 2014)
- Assignment #3 (due October 6, 2014)
- Assignment #4 (due October 15, 2014)
- Assignment #5 (due October 22, 2014)
- Assignment #6 (due November 5, 2014)
- Assignment #7 (due November 12, 2014)
- Assignment #8 (due November 19, 2014)
- Assignment #9 (due November 26, 2014)
- Assignment #10 - Data (due December 9, 2014)
Take Home (for graduate students only)
Classes, Exams, Grades, and References
- Lecture: Monday and Wednesday 16:00 - 17:30
- Location: HP 3349 (SunRay Lab)
- Office Hours: Monday and Wednesday 15:00 (or via appointment)
- Grading Policy for Undergraduate Students (Phys 4807):
- Homeworks = 50%
- Midterm = 20%
- Final = 30%
- Grading Policy for Graduate Students (Phys 5002):
- Homeworks = 40%
- Midterm = 20%
- Final = 30%
- Take Homes = 10%
- Textbook: Statistical Data Analysis, by Glen Cowan - Paperback 9780198501558 - Hardback 9780198501565.
- At the Carleton book store
or Oxford University Press.
See the book web page .
- Midterm date: TBD (open book: Cowan's text book only).
- Final exam date: TBD (open book: Cowan's text book only).
- Other Reference (used in previous years):
- Data Analysis: Statistical and Computational Methods for Scientists and Engineers, by S. Brandt.
Passing Condition
In order to pass the course, students should attempt all tests and assignments.
Missing tests and assignments must be accounted for, usually by bringing in a Doctor’s note, to the Instructor.
Students must obtain a minimum of 30 out of the 70 marks available for assignments and midterm. Term
work resulting in a mark less than this is "not satisfactory".
Students are expected to attend all lectures. A deferred final exam replaces only the final exam portion of the
marks and students must have completed satisfactory term work as explained above to be eligible. Deferred exams are generally
only granted to students who cannot take the regularly scheduled exam due to illness.
The following percentage equivalents apply to all final grades at Carleton:
A+ |
90-100 |
|
B+ |
77-79 |
A |
85-89 |
|
B |
73-76 |
A- |
80-84 |
|
B- |
70-72 |
|
|
|
|
|
C+ |
67-69 |
|
D+ |
57-59 |
C |
63-66 |
|
D |
53-56 |
C- |
60-62 |
|
D- |
50-52 |
|
|
|
|
|
|
|
|
|
|
F |
Failure (0-49). The grade of F is assigned when the student has failed to meet the conditions of “satisfactory” defined above. |
FND |
Failure with no deferred final examination allowed. The grade FND is assigned only when the student has failed the course on the basis of "not satisfactory" term work. |
Academic Accommodation
You may need special arrangements to meet your academic obligations during the term. For an accommodation request the processes are as follows:
Pregnancy obligation: write to me with any requests for academic accommodation during the first two weeks of class, or as soon as possible after the need for accommodation is known to exist. For more details see the Student Guide.
Religious obligation: write to me with any requests for academic accommodation during the first two weeks of class, or as soon as possible after the need for accommodation is known to exist. For more details see the Student Guide.
Academic Accommodations for Students with Disabilities: The Paul Menton Centre for Students with Disabilities (PMC) provides services to students with Learning Disabilities (LD), psychiatric/mental health disabilities, Attention Deficit Hyperactivity Disorder (ADHD), Autism Spectrum Disorders (ASD), chronic medical conditions, and impairments in mobility, hearing, and vision. If you have a disability requiring academic accommodations in this course, please contact PMC at 613-520-6608 or pmc@carleton.ca for a formal evaluation. If you are already registered with the PMC, contact your PMC coordinator to send me your Letter of Accommodation at the beginning of the term, and no later than two weeks before the first in-class scheduled test or exam requiring accommodation (if applicable). After requesting accommodation from PMC, meet with me to ensure accommodation arrangements are made. Please consult the PMC website for the deadline to request accommodations for the formally-scheduled exam (if applicable).
|
|