v 2.14.2: can't reproduce dec'ing example on p. 49 of SD2.12-part2 of the manual

[aliwang]

Dear SpinDynamica Team,

I am using version 2.14.2. (Outstanding program!)

When I run the example on page 49 of SD2.12-part2 of the manual, instead of getting decoupling for the first 0.1 s of the FID (as seen in the example), I get it for the last 0.1 s.

I cannot figure out why, and would be grateful for any insights into the problem. A notebook reproducing the example in the manual is attached to this post.

My best respects,
Andy

[MalcolmHLevitt]

Hi Andy, thanks for posting - I just checked the forum and picked up on your problem.

This puzzles me since when I add the line Needs["SpinDynamica`"] to the top of your file and run it, starting from a clean kernel, I do not get the behaviour you describe (see attached file). One possibility is that you were trying out some of the features in the documentation, such as the setting for $Chronology (end of part 1) and that a setting was thereby changed. Mathematica has the feature that anything you do in a session persists for the rest of the session, so maybe that is the problem.

Please retry, starting from a clean kernel (i.e. restart the kernel) and using the Needs line above to load SpinDynamica. Let me know how it goes.

All the best
malcolm

[aliwang]

Hello Malcolm,

Thank you for getting back to me on this question (and my chemical exchange question on a separate post).

I just did as you suggested, and attached a screenshot of the outcome. I started up the Mathematica program and typed the example from scratch into a new notebook.

I even re-downloaded your latest version of SpinDynamica.

Puzzling!
Andy

[MalcolmHLevitt]

Thanks Andy,
 this is very puzzling. Which version of Mathematica do you have?

 Can you run the attached file on your system including the Trace instructions and send the result back. If its too large to attach here please put it on dropbox and send me the link, or some other way. I want to see the complete file, not just screen clips.

 please start from a completely clean kernel and simply add a $Path instruction before Needs, if you need to do that.

 I will be away for a while now but I will look at this in the course of time. I havent seen a case where the same code gives different results when running the same version of Mathematica and SpinDynamica on separate systems.

 

[MalcolmHLevitt]

Incredible!

Since I was using a slightly older version of Mathematica (10.3) I decided to update to 10.4 and try again.

This time I got the same behaviour as you!

So there is a subtle difference between Mathematica 10.3 and 10.4 which causes different behaviour for this particular calculation, which is very unexpected. I hope to debug this after my summer break.

There should be no need for you to send me the files I requested in the previous post.

all the best
malcolm

[aliwang]

Hello Malcolm,

I am using version 10.4.1.0.

I am relieved you reproduced my observation. I was starting to think that I must be getting a bit soft in the head.

Enjoy your summer break,
Andy

[MalcolmHLevitt]

Hi Andy,
 while introducing SD v2.14.3, which is compatible with Mathematica 11, I also located and fixed this problem.
all the best
malcolm

[aliwang]

Excellent, Malcolm!

Thanks so much,
Andy