I need a guideline to learn MATLAB faster
Show older comments
I am a new learner of MATLAB kindly sir give me a guideline to learn MATLAB faster.
4 Comments
Jan
on 30 Jul 2011
"Faster" is questionable. MATLAB is extremly powerful, so learning to use it efficiently can take some years. But using the right techniques, you can get familiar with MATLAB in a week. And a lot of questions asked in this forum would be cleared implicitely. +1
Paulo Silva
on 30 Jul 2011
Jan it also depends heavily on your academic background, if you never done logic exercises or never tried to program in any language it would be very difficult to "learn MATLAB".
Jan
on 30 Jul 2011
@Paulo: You are right. The matrix oriented design of Matlab is easier to understand with a knowledge of vector algebra. It will take longer than a week to be familiar with the object oriented methods, if you never worked with OO before.
Matt Tearle
on 6 Jul 2016
Answers (6)
Paulo Silva
on 30 Jul 2011
1 vote
No warranty about learning faster but the tutorials do help, the documentation is awesome and you should use it and learn by yourself.
1 Comment
Jan
on 30 Jul 2011
The documentation has some limitations, e.g. MPOWER(scalar, matrix) is not explained sufficiently. Without deeper education in signal processing the Signal Processing Toolbox has a limited use only. The different filters are described well, but if you do not know, what you are searching for, there is no chance to solve a scientific problem. Nevertheless, it was not the interntion of the documentation to replace WikiPedia and the necessary lessons at the university. Therefore I agree with your estimation "awsome". Most of all if the M-source codes of the toolbox functions are taken as advanced examples. +1
Image Analyst
on 30 Jul 2011
1 vote
Found this on the Link Exchange. http://www.mathworks.com/matlabcentral/linkexchange/links/2920-learn-matlab-in-30-minutes Maybe you might give it a shot.
7 Comments
Jan
on 30 Jul 2011
30 minutes - smashing. But to avoid frustration of the OP: You will have to invest more minutes to build a professional MATLAB skill.
Paulo Silva
on 30 Jul 2011
30 minutes, that's just enough time to start MATLAB, open the documentation, look at the documentation topics and close MATLAB
Jan
on 30 Jul 2011
@Paulo: I've measured your test suite: Matlab 2009a, 1.5GHz Pentium-M, WinXP. I've read the 1st level topics of Release Notes, MATLAB, IPT, MATLAB compiler, SPT. Elapsed time: 71 sec.
Reading the 2nd level topics also took about the double time.
Do you mean reading the topics in all levels?
Paulo Silva
on 30 Jul 2011
Jan that was just my guess, no test suite, ok maybe reading all levels + any installed toolboxes...
Jan
on 30 Jul 2011
@Paulo: Sorry, I forgot a ;-) . I'm simply to embarrased by the MATLAB-in-30-minutes idea.
@IA: The quality of the text can be improved. Some sentences will confuse the beginners ("!operaot (Not Operator) will inverse 0 with 1 and 1 with 0 in any value." - really?). but it is a <30 minutes crash course. +1
Image Analyst
on 31 Jul 2011
Jan: I have no idea - I didn't actually try the link myself, I just threw it out there for whatever it's worth (which apparently is not much). Honestly, I have yet to see code that's as explicit and well documented as my own, and the quality of what's posted in the File Exchange is usually way below my standards. The authors' algorithms may be good but almost always they write cryptic, hard -to-follow code with few if any comments. I'm sure you'd agree. Apparently it does take more than 30 minutes to learn, and some people never learn to write robust, maintainable, well documented code even after decades. I hope none of those people ever go to work for Boeing or Airbus, or a medical equipment manufacturer, or a nuclear power plant.
Jan
on 31 Jul 2011
@Image Analyst: You do know, that programmers with limited skill are working in sensitive areas. They let the ArianeV explode by confusing DOUBLE values and FLOAT overflow values - an epic 2.0e9 $ fault. A German health insurance had developped a new software, which solved all tests for a data set of 100 patients. To the big surprise of the managers it was absolutely impossible to expand the data base to 5.000.000 patients, because the fundamental concept resulted in an exponential growth of the latency. 25e6 $ burnt.
The list of dilettantish software tragedies could fill some books, but I stop here. My conclusion is, that a software can never replace savvy and common sense.
Jan
on 30 Jul 2011
1 vote
- Read the "Getting Started" chapters in the documentation. They are designed to assist newcomers.
- Read the help and doc texts whenever you have difficulties with a specific command.
- Use the debugger if problems occur. See "doc debug".
- Read the error messages twice. They contain important information
- Avoid common pitfalls: Read the FAQ before the problems have a chance to occur. Most of all:
- Do not use names, which contain indices like A1, A2, A3, ... Better use e.g. a cell array A{1}, A{2}, A{3}, because you can access it using an index.
- You do not need EVAL - it produces just troubles. The only exception is getting numerical values for symbolic expressions. But for all other cases: Use a better EVAL-free solution.
- Do not expect 0.3 - 0.2 - 0.1 to be zero. Floating point values have a limited precision, such that -2.776e-17 is not wrong.
Andrew
on 30 Jul 2011
0 votes
deep
on 30 Jul 2011
0 votes
where is the documentation..??
2 Comments
Paulo Silva
on 30 Jul 2011
http://www.mathworks.com/help/techdoc/
Jan
on 30 Jul 2011
@deep: And locally on your computer, typing "help" in the command window shows a short help, "doc" opens the full documentation. Try "help sin", "doc plot".
Matt Tearle
on 6 Jul 2016
0 votes
You can now get started in MATLAB by taking the free MATLAB Onramp. It teaches you the basics of MATLAB interactively -- you actually try out the MATLAB code and you see the results and feedback to help you
Categories
Find more on Matrix Indexing in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!