3
$\begingroup$

I am doing some complicated and tedious calculation on iterated commutators. A typical term in my calculation looks like $$[x_a,[[[x_b,x_c]-x_d,x_e],[x_f,x_g]]]\text{.}$$

(I am considering commutators on Lie algebras so the minus sign $-$ is well defined here.) In the notation above there is no ambiguity; but when I have dozens of terms like that it looks horrible and a mistake in calculation is almost inevitable. I have simplified the notation a little bit by forgetting the $x$'s and write the commutator as $$[a,[[[b,c]-d,e],[f,g]]]\text{,}$$ but it still looks confusing. Not sure if I should continue to simplify it and simply write $$[a[[([bc]-d)e][fg]]]$$ instead. I tried to use physical papers as my draft papers and also opened a rich text document to note down the calculation; but both ways look awkward and it is difficult for me to "see" when should I cancel two terms (under some conditions) or use the Jacobi's identity.

So my question is:

Is there any "smart" notation that I can use for calculation like this and make the calculation less fallible?

I am not referring to writing for publishing, but just for calculation on one's own draft paper.

$\endgroup$
  • $\begingroup$ I wonder if it would be difficult to try to get the computer to simplify expressions of this type. $\endgroup$ – MJD Dec 12 '14 at 15:55
  • $\begingroup$ @MJD, are you referring to using some softwares such as GAP? $\endgroup$ – Zuriel Dec 12 '14 at 15:56
  • $\begingroup$ How about $[x_a,x_b]=ab$? You'll still have to keep track of parentheses, but I think it uses about as little notation as possible. $\endgroup$ – user134824 Dec 12 '14 at 15:57
  • $\begingroup$ I suggest to use larger brackets. $\big[a,[b,c]\big]$ $\endgroup$ – Martin Brandenburg Dec 12 '14 at 15:58
  • 1
    $\begingroup$ @user134824, sounds good! I am also considering using different colours to keep track of parentheses; if I do so, I thinking using pens/pencils should be easier than using LaTex/Microsoft Word. $\endgroup$ – Zuriel Dec 12 '14 at 16:01
1
$\begingroup$

I think that you're going to have some cluttered-looking formulas no matter what.

However, instead of $[X, Y]$, one could write, say, $\mathcal{L}_X(Y)$.

You could obviously even shorten this to $X(Y)$, like function notation. If you then switch to exponent notation for functions (as, $f(x) = x^f$), and especially if you're willing to mix these notations, then (assuming I parsed it right!) your sample formula could become, for instance: $$ ([f, g]^{[[b, c] - d, e]})^a $$ which to my eye is slightly more readable than what you had.

Obviously this could cause a bunch of other problems for you! For instance: (1) you may want to reserve exponential notation for, well, exponents. (2) Visually, it switched the order. So even though $a$ started out way on the left, now it's way on the right.

We could perhaps try to lessen the confusion in both (1) and (2) at once: Write ${}^xy = [x, y]$, so now your formula is $$ {}^a({}^{[[b, c] - d, e]}[f, g]) $$ You could also use subscripts instead of superscripts, etc. It's a matter of taste whether something like this is preferable. Obviously, do whatever works best for you!

$\endgroup$
  • $\begingroup$ Thank you for the suggestion! So you mean we write ${}^xy=[x,y]$ sometimes but not always, right? Since otherwise the notation will look like $${}^a({}^{^{{}^bc-d}e}({}^fg)).$$ And when to write ${}^xy$ and when to write $[x,y]$, there is no rule; right? $\endgroup$ – Zuriel Dec 12 '14 at 16:26
  • $\begingroup$ Yeah, my suggestion would be to mix notations, assuming you're comfortable with that. So you could still use the $[x,y]$ notation sometimes. There wouldn't be have to be any rule for when to use each notation; you'd just use whatever you thought looked more clear. (Like how sometimes people write $2*3$ and sometimes people write $(2)(3)$; either is acceptable, and you just use whatever is most convenient at the time). $\endgroup$ – mathmandan Dec 12 '14 at 16:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.