2016: Difference between revisions

From Advanced Labs Wiki
Jump to navigation Jump to search
No edit summary
Line 37: Line 37:
'''Safety.''' Use your common sense in all situations. In these labs you'll encounter manageable hazards. Follow the instructions carefully. '''Food and drink are not allowed near the labs.''' Safety also follows from orderliness: please keep (and leave) the lab in an organized state. '''When in doubt ask an [[#Instructors|instructor]].'''
'''Safety.''' Use your common sense in all situations. In these labs you'll encounter manageable hazards. Follow the instructions carefully. '''Food and drink are not allowed near the labs.''' Safety also follows from orderliness: please keep (and leave) the lab in an organized state. '''When in doubt ask an [[#Instructors|instructor]].'''


==Analysis==
===Analysis===


Analysis is the evaluation of data towards an interpretation that accounts for errors in the data. Roughly speaking there will be three analysis steps in this class:  
Analysis is the evaluation of data towards an interpretation that accounts for errors in the data. Roughly speaking there will be three analysis steps in this class:  
Line 45: Line 45:
#Assess the impact of systematic errors.
#Assess the impact of systematic errors.


===Tool for Analysis===
The tool we will use for the analysis is the Python programming language and its numerical and scientific computing modules. Specifically you will use IPython Notebook.


The tool we will use for the analysis is the Python programming language and its numerical and scientific computing modules.
===Reports===
 
You will submit an IPython Notebook with your [[Report|report]].
 
==Reports==
   
   
For each experiment you will write a report.  
For each experiment you will write a report. The reports will be presented in a standard scientific format with use of figures and tables. The format should have 1" margins with no smaller than 11 point font. The maximum number of pages is 6, including figures and tables. The quality of your scientific writing and other presentation elements is crucial. See the [[media:Report_Checklist.pdf|report checklist]] for key pointers when writing your reports.
 
The reports will be presented in a standard scientific format with use of figures and tables.
 
The format should have 1" margins with no smaller than 11 point font. The maximum number of pages is 6, including figures and tables.
 
More information can be found in the [[media:Report_Checklist.pdf|report checklist]].
 
===Feedback===
 
For the first two reports, students will meet one-on-one with the professor to discuss their graded work.
 
===Tool for Reports===
 
The document preparation system for reports is [http://latex-project.org/ LaTeX].
 
You can also download LaTeX freeware for your personal computers (e.g., TeXworks on all platforms, TeXShop for Mac, TeXnicCenter for Windows).
 
Good online LaTeX editors also exist (e.g., ShareLaTeX, papeeria).
 
The computers in the PUC lab have various installations of LaTeX editors/compilers.
 
==Readings==
 
Lectures will be based on


'''Bevington & Robinson, ''Data Reduction and Error Analysis for the Physical Sciences'', 3rd Edition, McGraw-Hill, ISBN 0-07-247227-8, 2003'''
'''LaTeX.''' The document preparation system for reports is [http://latex-project.org/ LaTeX]. You can also download LaTeX freeware for your personal computers (e.g., TeXworks on all platforms, TeXShop for Mac, TeXnicCenter for Windows). Good online LaTeX editors also exist (e.g., ShareLaTeX, papeeria). The computers in the PUC lab have various installations of LaTeX editors/compilers.


I will have a couple class copies that you can read in the lab. '''Please keep these in the PUCLab.'''


==Schedule==
==Schedule==
Line 169: Line 140:
The grade of late work will be multiplied by exp(-(days late)/7), where days late can be fractional (starting from midnight).  
The grade of late work will be multiplied by exp(-(days late)/7), where days late can be fractional (starting from midnight).  


<!--


===Anatomy of an Experiment===


'''Experiment Execution.''' The first step in executing an experiment is to have a good idea of the phenomenon being measured -- the reason why you're doing the experiment. Then you need to have a thorough knowledge of the experimental apparatus. With this preparation you will be able to take data. But obtaining measured values is not enough. You need both values and errors. You need to conduct the experiment in a way that estimates systematic errors and statistical errors. Systematic errors can be checked for by conducting the experiment in more than one way that should, e.g., give the same result and checking for discrepancies. Statistical errors may be obtained by repeating the experiment and evaluating the sample variance of the data or there might be an analytic expectation for the statistical error, as in the case of counting experiments.
==Other Resources==


'''Data Analysis and Interpretation.''' The input to data analysis consists of measured values and their errors. You then fit this data with some physical model. If the fit is "good", then you can believe  the best-fit model parameters and associated model parameter errors. These model parameters tell you something about the physical world.
'''Data Analysis'''
==Readings==


'''Presentation''' Lab reports constitute the language of the course. The sections of a report are
Lectures on data analysis are loosely based on the text
*Abstract -- Summarily say the aim of the experiment and what you used to measure the phenomenon. Then quote your result which is usually some physical parameter ''with errors''.
*Introduction -- Describe ''qualitatively'' the phenomenon being measured and the history of the measurements and theory behind the current experiment (seminal works cited etc). This should not contain much information about what you did in the experiment-- just roundly what you aim to do. The intro is mainly useful background and can be relatively brief.
*Theory -- Introduce ''quantitatively'' the physical effect that you're trying to probe. Introduce equations.
*Experiment Description and Data -- Describe the experiment setup and procedure. Also describe the data and errors. In particular you'll likely be giving the averages of your many samples collected and errors on those average. These should appear in a plot and, when possible, a table.
*Data Analysis -- Derive a theoretical interpretation from the data propagating errors  to theoretical model parameters etc. If appropriate, discuss the "goodness" of the model fit.
*Discussion -- Interpret your results and discuss what may have gone wrong if, e.g., the fit in the Data Analysis section was not good.
*Conclusion -- A short section where you summarize the paper. This section could possibly include future directions to take the work.


===Lab Report Specifications===
Bevington & Robinson, ''Data Reduction and Error Analysis for the Physical Sciences'', 3rd Edition, McGraw-Hill, ISBN 0-07-247227-8, 2003
The reports are to be created on a computer with computer generated graphics, plots, etc. The document preparation system for the reports is LaTeX. The computers in the PUC lab have various installations of LaTeX editors/compilers. You can also download freeware for your personal computers. I think the online editor "ShareLaTeX" is pretty good.


The lab reports should have an abstract, an introduction, a theory section,  description of the experiment (apparatus and procedure) ''and reduced data and errors'', description of the analysis of the data, discussion of results, a conclusion, and a bibliography.  
We have a couple class copies that you can read in the lab. Please keep these in the PUCLab.


The format should have 1" margins with no smaller than 11 point font. The maximum number of pages is 6, including figures and tables. Be concise.
Other Data analysis texts include


A standard strategy is to create your figures first in order to guide the body of the text.
-->
==Other Resources==
'''Data Analysis'''
*Press, Teukolsky, Vetterling, Flannery, ''Numerical Recipes in C'' (Available online)
*Press, Teukolsky, Vetterling, Flannery, ''Numerical Recipes in C'' (Available online)
*Lupton, "Statistics in Theory and Practice"
*Lupton, "Statistics in Theory and Practice"

Revision as of 21:11, 6 January 2016

Welcome to Advanced Physics Lab 2016!

Instructors

Professor: Tobias Marriage (marriage@jhu.edu)

Teaching Assistant: Devin Crichton (dcrichton@jhu.edu)

Lab Guru: Steve Wonnell (wonnell@pha.jhu.edu)

Classes

The class times are Monday 10:00-12:50 and 1:30-4:20.

The class location is the Physics Undergraduate Computer lab (PUClab). A PUCLab login for the computers may also be useful. Access to the lab and computers is managed by Steve Wonnell.

Laptops should be brought to class.

The first half of class will be devoted to lecture and discussion. The first half of the 10am and 1:30am classes will be similar.

The second half of class will be devoted to work with hands-on help from instructors.


Essential Elements

In this class, you will conduct experiments, analyze data, and write up your results in reports.

Experiments

Brownian Motion (BM): The goal of this experiment is to estimate the Boltzmann constant using a measurement of the Brownian motion of microscopic spheres.

Speed of Light (SoL): In this experiment you use the classic "time of flight" measurement by Foucault to estimate the speed of light.

Galactic Rotation (GR): In this experiment you'll use a radio telescope to measure the rotational velocities in the Galactic disk and judge whether the data better fits a model with or without dark matter.


Safety. Use your common sense in all situations. In these labs you'll encounter manageable hazards. Follow the instructions carefully. Food and drink are not allowed near the labs. Safety also follows from orderliness: please keep (and leave) the lab in an organized state. When in doubt ask an instructor.

Analysis

Analysis is the evaluation of data towards an interpretation that accounts for errors in the data. Roughly speaking there will be three analysis steps in this class:

  1. Process the raw data sets into a reduced dataset with errors,
  2. Interpret the reduced data in the context of a physical model, finding best estimates and errors on model parameters, and
  3. Assess the impact of systematic errors.

The tool we will use for the analysis is the Python programming language and its numerical and scientific computing modules. Specifically you will use IPython Notebook.

Reports

For each experiment you will write a report. The reports will be presented in a standard scientific format with use of figures and tables. The format should have 1" margins with no smaller than 11 point font. The maximum number of pages is 6, including figures and tables. The quality of your scientific writing and other presentation elements is crucial. See the report checklist for key pointers when writing your reports.

LaTeX. The document preparation system for reports is LaTeX. You can also download LaTeX freeware for your personal computers (e.g., TeXworks on all platforms, TeXShop for Mac, TeXnicCenter for Windows). Good online LaTeX editors also exist (e.g., ShareLaTeX, papeeria). The computers in the PUC lab have various installations of LaTeX editors/compilers.


Schedule

Date 1st Half of Class Reading 2nd Half of Class and Homework
Jan 25 Class Overview; Experiments, Measurement & Errors 1 Bev. Ch1 Install Ipython Notebook; Install LaTeX or use On-line LaTeX editor.
Media:LaTeX_Example.txt, Python_Example
Feb 01 Measurement and Errors 2; BM Introduction; LaTeX Tutorial Bev. Ch 1&2 Start BM
Feb 08 Probability Distributions; BM Analysis Discussion; Python Tutorial 1 Bev. Ch 2 Start analysis & writing initial sections of report
Feb 15 Propagation of Errors Bev. Ch 3 Working on Full Report
Feb 22 Method of Maximum Likelihood Bev. Ch 4 BM Work Due Friday Feb 26 5pm (Email TA)
Feb 29 Student's t and Chi-sq Distributions; SoL Introduction Bev. Ch 6 Start SoL; Schedule BM Report Reviews
Mar 7 Linear Least Squares 1; SoL Analysis Discussion; Python Tutorial 2 Bev. Ch 6 BM Report Reviews
Spring Break
Mar 22 Review Line Fitting Bev. Ch 7 BM Report Reviews
Mar 28 General Linear Least Squares SoL Report Due Friday April 1 5pm (Email TA)
Apr 4 Nonlinear Fitting 1; GR Introduction Bev. Ch 8 Start GR; Schedule SoL Report Reviews
Apr 11 Nonlinear Fitting 2; Python Tutorial 3 Bev. Ch 8 SoL Work Reviews
Apr 18 Experiment/Analysis Topic Review - SoL Report Reviews
Apr 25 Reserved for Overflow - GR Work Due Friday May 1 5pm (Email TA)

Grading

Grades breakdown as 1/3 experiment execution, 1/3 data analysis, 1/3 scientific writing.

Collaboration Policy

Students are encouraged to discuss experiments, analysis, and other course related issues with the instructors and their peers. However, each person should obtain their own data, perform their own data analysis (e.g., no code sharing), produce their own plots, and write their own report. For SoL, two students will need to collaborate, but each should obtain their own dataset.


Ethics

The strength of the university depends on academic and personal integrity. In this course, you must be honest and truthful. Ethical violations include cheating on exams, plagiarism, reuse of assignments, improper use of the Internet and electronic devices, unauthorized collaboration, alteration of graded assignments, forgery and falsification, lying, facilitating academic dishonesty, and unfair competition. For more info: http://e-catalog.jhu.edu/undergrad-students/student-life-policies/.

Work Submission and Late Work

Submitting Reports and Notebooks

Reports should be submitted in PDF format to the professor through email together with iPython Notebooks containing the analysis.

Submitting Raw Data

Initial submission of the raw data will be done through google drive or email depending on the lab.

If your initial dataset is flawed, you will have the opportunity to retake and resubmit it for partial credit.

Late Policy

The grade of late work will be multiplied by exp(-(days late)/7), where days late can be fractional (starting from midnight).


Other Resources

Data Analysis

Readings

Lectures on data analysis are loosely based on the text

Bevington & Robinson, Data Reduction and Error Analysis for the Physical Sciences, 3rd Edition, McGraw-Hill, ISBN 0-07-247227-8, 2003

We have a couple class copies that you can read in the lab. Please keep these in the PUCLab.

Other Data analysis texts include

  • Press, Teukolsky, Vetterling, Flannery, Numerical Recipes in C (Available online)
  • Lupton, "Statistics in Theory and Practice"

LaTeX

Previous Year's Tutorials (No longer supported -- use at own risk!)

And of course... Wikipedia!

Retrieved from "https://advlabwiki.johnshopkins.edu/index.php?title=2016&oldid=1383"