1
$\begingroup$

It might be a silly question. Actually I'm facing a problem expanding $(1+\frac{2}{x})^{\frac{i}{2}}$ for small $x$. Mathematica can not expand it. But it can of course expand $(1+\frac{2}{x})^{\frac{1}{2}}$ for small $x$. Is it related to properties of the expression ( like convergence etc) or I need to use mathematica a bit carefully?

$\endgroup$
  • $\begingroup$ Is i the imaginary unit in your expression? $\endgroup$ – march Oct 22 '15 at 22:59
  • 1
    $\begingroup$ You can try raising the result of expanding the square root to I... $\endgroup$ – J. M.'s technical difficulties Oct 22 '15 at 23:00
  • $\begingroup$ @march yes. It's $\sqrt{-1}$ $\endgroup$ – Physics Moron Oct 22 '15 at 23:14
  • $\begingroup$ @J.M. Actually I tried that. But didn't get anything useful. $\endgroup$ – Physics Moron Oct 22 '15 at 23:24
4
$\begingroup$

Perhaps this, from J.M.'s hint:

Series[(1 + 1/x)^(1/2), {x, 0, 3}]
Series[(Normal@%)^I // Expand, {x, 0, 3}]
Expand@Normal@%

enter image description here

Or, perhaps this: (per another of J.M.'s hints):

Series[(1 + 1/x)^(1/2), {x, 0, 3}]
Series[%, {x, 0, 3}]^I
Normal@Expand@%

enter image description here

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thanks! It looks good. But probably I should check it by hand once.. $\endgroup$ – Physics Moron Oct 22 '15 at 23:41
  • 3
    $\begingroup$ Actually, Series[(1 + 1/x)^(1/2), {x, 0, 3}]^I // Normal is what I'd have done. ;) $\endgroup$ – J. M.'s technical difficulties Oct 23 '15 at 0:14
  • $\begingroup$ @J.M. Well of course you'd have done something smarter than me. :) $\endgroup$ – march Oct 23 '15 at 1:18
1
$\begingroup$

Is this acceptable where you introduce the factor "c"?

Series[(1 + 2/x)^(I/3), {x, c, 4}]

(* (1 + 2/c)^(I/3) - (2 I (1 + 2/c)^(-1 + I/3) (x - c))/(3 c^2) + (
 2 I (1 + 2/c)^(I/3) ((3 + I) + 3 c) (x - c)^2)/(9 c^2 (2 + c)^2) - (
 2 I (1 + 2/c)^(I/
  3) ((34 + 18 I) + (54 + 18 I) c + 27 c^2) (x - c)^3)/(
 81 c^3 (2 + c)^3) + ....*)  

Limit[%, x -> 0]

(* (((2 + c)/
  c)^(I/3) ((3260 + 
     2640 I) + (6912 + 4704 I) c + (5508 + 2916 I) c^2 + (1944 + 
      648 I) c^3 + 243 c^4))/(243 (2 + c)^4) *)
| improve this answer | |
$\endgroup$
  • $\begingroup$ Then I need to replace c by x? $\endgroup$ – Physics Moron Oct 22 '15 at 23:35
  • $\begingroup$ @pinu The meaning of "c" here is the point at which you expand the series on "x". So you reach singularity if you put "c=0", maybe try a small number for c. $\endgroup$ – thils Oct 22 '15 at 23:38

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.