Practice With Matrices and Eigenvalues/Vectors - Griffiths Quantum, Problem A.28
In this video we work through parts A, B, and C of Griffiths Quantum Mechanics problem A.28. In this problem we practice working with matrices, eigenvalues, and eigenvectors.
All code for this episode can be found on the following GitHub page.
GitHub Repo Directory: https://github.com/nkphysics/Computational-Linear-Algebra-/tree/master/7_Cholesky
In this 7th video in this computational linear algebra series we cover a higher level variant of the LU Decomposition called the Choleksy Decomposition. The Choleksy decomposition's key benefit is that there we will only need to keep track of one lower triangular matrix, generally referred to as the Cholesky Factor. If we need to get back to our original matrix we can just multiply our lower triangular matrix (Cholesky Factor) by its transpose. By keeping track on only one matrix (the Choleksy Factor) we greatly improve out performance computationally.
The one key initial condition which allows us to do a Cholesky decomposition though is that our matrix we want to decompose must be considered "positive definite."
For more information on positive definiteness please see the following online resources:
MIT Mini Lecture #1 : https://youtu.be/ojUQk_GNQbQ
MIT Lecture: https://youtu.be/xsP-S7yKaRA
MIT Mini Lecture #2: https://youtu.be/lpnY5QVjU5w
Quick Tutorial on checking positive definiteness: https://youtu.be/ttMZB5Gm_fM
All credit goes to Griffiths for writing a fantastic quantum mechanics textbook. In this video we work through part A of problem 1.5 out of Griffiths Quantum mechanics.
This is just a quick discussion I wanted to have with all of my viewers were I answer some of your questions and comments. I mainly answer the questions "How did you learn NICER data processing?"
I still cannot find the link to the NICER help desk email. If/when I find that email I will post is in a pinned comment.
Find the recorded NICER Workshop videos here:
https://heasarc.gsfc.nasa.gov/docs/nicer/data_analysis/workshops/index.html#spring21
I decided at the last minute not to include my twitter page so I am not mixing streams.
Here is my github: https://github.com/nkphysics
AutoNICER is a bit of software I developed to automate the procedure of retrieving and reducing data gathered with the NICER mission. This video highlights the functionality of autoNICER and how to use it.
autoNICER on GitHub: https://github.com/nkphysics/autoNICER
autoNICER on PyPI: https://pypi.org/project/autonicer/
Contents:
00:00 - Intro
02:30 - Evidence
04:33 - GitHub Repository
04:45 - Prerequisites
05:50 - Installation
06:34 - HEASarc Archive
07:00 - Running autoNICER
09:24 - Selecting Datasets
13:45 - Retrieving and Reducing Data
15:57 - Details on Data Reduction commands
17:26 - Summary of run
18:24 - Output Log
19:55 - Contributing
21:23 - Future Plans
23:45 - Where I need help
26:04 - Conclusion
DISCLAIMER:
I (the creator of this video) am not affiliated with NASA, any space flight center, any NASA mission or any university or research institute. I am an independent researcher who utilizes HEASoft and the NICER mission's archival data to facilitate my own independent research into neutron stars.
Visit NICER at HEASARC: https://heasarc.gsfc.nasa.gov/docs/nicer/
Check out HEASoft: https://heasarc.gsfc.nasa.gov/docs/nicer/
NICER is one of the newer missions within HEASARC and in order to analyze and of the data successfully you must process the data first. This is a video tutorial showing how to process NICER data with the nicerl2 tool within heasoft.
See HEASARC instructions here:
https://heasarc.gsfc.nasa.gov/docs/nicer/data_analysis/nicer_analysis_guide.html
If you have any issues using nicerl2 pleas leave a comment and I will answer as fast as possible.
This video is the first of many in my series on the topic of Computational Linear Algebra. This is the introductory video so it is all about the linear system. Linear systems are the backbone of linear algebra so many may already be familiar with Linear Systems. In this video we walk through very briefly how to solve a linear system by hand using Gaussian Elimination. Then we dive into two computational examples using the octave and python programming languages. Understanding how to solve a linear system computationally in one of these languages will be crucial in further understanding higher level concepts in Computational Linear Algebra.
The code for these two examples can be found here:
Github:
https://github.com/nkphysics/Computational-Linear-Algebra-/tree/master/1_Introduction
Gitlab:
https://gitlab.com/n_space_cowboy/computational-linear-algebra/-/tree/master/1_Introduction
Chapters:
0:00 - Intro
1:00 - Preface (You can skip this if you aren't interested in my goals for the series of Computational Linear Algebra)
2:13 - Solving by hand with Gaussian Elimination
3:30 - Ex 1, Solving a 10x10 system
7:00- Ex 2, Solving a 1000x1000 system
9:20 - Closing Remarks
Code: https://github.com/nkphysics/Computational-Linear-Algebra-/tree/master/10-Least_Squares
numpy.linalg.lstsq documentation: https://numpy.org/doc/stable/reference/generated/numpy.linalg.lstsq.html
Table of Contents:
0:00 - Introduction
0:58 - What causes the problem with an over-determined system
3:22 - The Method of Least Squares
5:01 - Method of Least Squares without Numpy
8:42 - Method of Least Squares with Numpy
In this video we discuss how to find an approximate solution to an over-determined system (a linear system with more rows than columns). We can find an approximate solution to an over-determined system with a little bit of knowledge of sub-spaces and projections. We obtain an approximation to an over-determined system by using a method called "The Method of Least Squares" which plays a crucial role on the back end of may routines involving machine learning, AI, and model fitting. This approximation is often referred to as a "Least Squares Solution".