MANE 3351
Lecture 2, August 25
Classroom Management
Agenda
- Attendance
- Textbook
Resources
Handouts
Calendar
| Lecture | Date | Content |
|---|---|---|
| 1 | | Welcome to Class, Syllabus | |
| 2 | August 25 | Error Analysis |
Assignments
- Get textbook
Lecture 2 Content
Embrace Your Inner Skeptic
Definition: Numerical Methods
Embrace your Inner Skeptic
Evaluating Numerical Methods
Numerical methods should consider these two evaluation criterion :
- Accuracy (or error) analysis
- Speed
Introduction to Error Analysis
Error Measurements
Let \(v_A\) be the approximate value and \(v_E\) be the exact value
- Absolute error: \(|v_A-v_E|\)
- Relative error: \(\frac{|v_A-v_E|}{v_E}\)
- Percentage error: \(\frac{|v_A-v_E|}{v_E}\times 100\%\)
Sources of Error
Sources of Errors
- Round-off errors are due to the fact that the computers represent numbers in a finite number of bits and bytes
- Truncation errors are errors that emerge from the approximation of the mathematical model
- Model errors are due to the fact that the mathematical model usually is an approximation of the physical reality
Number Types in Python
Python supports int, float, and complex.
- Integers can be represented exactly
- Float and complex variables can be represented approximately
Computer Implementation of Floats
IEEE Standard 754 Floating Point Numbers
Floating Point Representation
Perils of Floating Point
Single Precision Range
Single Floating-Point Precision
-
"numerical methods." A Dictionary of Computing.. Retrieved August 27, 2019 from Encyclopedia.com: https://www.encyclopedia.com/computing/dictionaries-thesauruses-pictures-and-press-releases/numerical-methods ↩
-
Cheney and Kincaid (2004), Numerical Mathematics and Computing, 5th edition. ↩
-
Geeks for Geeks. Retrieved August 28, 2019 from: https://www.geeksforgeeks.org/ieee-standard-754-floating-point-numbers/ ↩
-
FloatConvertor. Retrieved August 28, 2019 from: https://www.h-schmidt.net/FloatConverter/IEEE754.html ↩
-
Perils of Floating Point. Retrieved August 28, 2019 from: http://www.lahey.com/float.htm ↩
-
IEEE 754 Floating Point Representation. Retrieved August 28, 2019 from: http://cs.boisestate.edu/~alark/cs354/lectures/ieee754.pdf ↩↩






