on the wp34s, asin (8+9i) gives 0.84+3.18i. Similar calculation on 50g and 42s gives 0.725+3.18i. Ln (2+3i) gives 1.28+0.58i on wp34s but on 50g and 42s gives 1.28+0.982i... Did i do something wrong?

*Edited: 30 Aug 2012, 4:04 a.m. *

A problem with complex LN I suspect.

- Pauli

new fixed firmware soon for complex functions? :(

Curiously, I cannot replicate your results on my wp34s (version 3225). I get the same results as indicated for the 42s and 50G.

Hi, I get the same results as 42s and 50G, with Ver 3233.

Regards

John

[Edit] Just figured out the problem! You are entering the values in the wrong order. Enter the imaginary part first then the real part.

*Edited: 30 Aug 2012, 4:52 a.m. *

Not a problem with LN. More like an entry problem.

3 ENTER 2 CPX LN gives the right answer.

2 ENTER 3 CPX LN gives your answer.

The complex component goes in Y not X.

- Pauli

yes, that is it. it's opposite to complex number entry on 42s. However, the stack entries are consistent for both the initial argument and result, i.e. ST X is the real part and ST Y is the imaginary part. Hence for entry, imaginary part is entered first.

Thanks

*Edited: 30 Aug 2012, 5:09 a.m. *

:-)

This is a typical example. Either Real^Imaginary or Imaginary^Real is the most basic notion needed to use a complex-capable calculator. One can't assume all models do it the same way, so a quick check (either reading the manual or doing a simple operation to confirm) is in order.

For instance, the 41Z does it either way, depending on the preferred method:

- Im^Re if using the complex stack

- Re^Im if via the "^IM/AG" function.

Cheers,

ÁM.

*Edited: 30 Aug 2012, 5:12 a.m. *

That's why I greatly prefer RPL systems from RPN system (even highly sophisticated RPN system such as WPS34S); one object per stack level and one stack level par objet when or ever it's dealing with complex or very complex objects or not!

well, i guess given the limitation of the 30b platform and stack that's the way it works.

can't wait to play around some more ;)

Be assured we were most willing to do this for the WP 34S the same way, alas, the hardware (i.e. the space provided) doesn't allow for handling such objects the way the HP-42S did it. And "space provided" isn't meaning that ... ummh ... suboptimal display alone :-( but also the amount of RAM available for storing such object types. Give us a reasonable LCD and we can at least display a complex number completely - else there's simply no way.

I think i said it. It's a rare thing nowaday to see non graphic calculator with dot matrix multiline instead of segmented multiline display nowaday. It's also nice to have the customable soft menu possible with such thing... Among other benefits..

Well, as it happens on the 42S each stack level contains the "complete" complex number, or even a Matrix - so no need to have RPL for that.