Unfortunately, I only have Express, so I can't help with the presentation of the symbolic result.
Perhaps, try putting the variables in a vector?
[x_1]
solve, |y_1|
[z_1]
You can assign the result to a variable and transpose it.
However, is there any particular reason you want to use the symbolic processor to solve the equation, when a numeric solution seems to give reasonable results?
Stuart
Stuart's got it: simply assign result to a vector and then take transpose. P11 file attached.
The symbolic solve will always provide the solutions row-wise and there is no option to set which will change this behaviour. You will get further rows for non-linear systems and each row represents a solution set.
I also think that in case you are just looking for the numeric values its better to use a numeric method and not the symbolics. Stuart has already shown some ways. Other options are the usage of the "lsolve" function
or using a solve block with "find" (you have to provide arbitrary guess values) which displays its results column-wise as you like it.
Using the symbolics makes sense if you are looking for symbolic solutions (in case your system contains some variables)
and, yes, in this case displaying the results column-wise sure makes sense 😉
Another reason for using the symbolics could be that you want 'exact' results, displayed in fractions (and maybe pi, e, etc) and not a decimal approximation. Unfortunately the symbolics automatically switches into float-mode as soon as just one decimal point is present. So you must also provide your information with fractions, i.e. 3/2, even though 1.5 is also an exact value.
You may also assign the result of the symbolic "solve" a row vector (1 x 3 matrix) with three variables which you then can evaluate symbolically or numerically and use for further calculations.
EDIT: If you really want to use the symbolic engine and insist on a column vector right from the start, there is a way to do it: You have to use a solve block with "find" which is evaluated symbolically:
@StuartBruff wrote:
Will the symbolic processor accept a vector [x,y,z] as a parameter to the solve keyword?
(where [x,y,z] is a column vector)
Yes, the symbolics accepts the list of variables as I did, but also a vector. It does not matter if its a column or a row vector.
It also accepts a mixture
@Werner_E wrote:
I understand and a list type sure would be very welcome as would an empty array, as would ...
But maybe the highlight of the next version is just a user definable border around regions - what a masterpiece!
Hey, I know it might seem like taking imagination to Verrücktheit levels, but don't knock it.
One look at that, and a Company Board couldn't fail to get behind an engineer's request for such an upgrade, and throw their financial might behind their appeals to the Mathcad Prime 13 Requirements Team (and, perhaps, allow an enterprising engineer the opportunity to slip in a few minor requests alongside it (such as empty arrays or lists)).
Stuart
@Werner_E wrote:
Yes! That's quite a frame!
But you didn't mention multidimensional arrays or the implementation of cuneiform as a minor request—should I be concerned?
I thought that was taken for granted? Lists haven't been mentioned much (or at least not by that name), so if I don't mention MDAs, there's a very slight chance the Committee for Denying Stuff for Bruff might miss calls for their implementation.
However, you do raise a good point. I note from the Prime 12 presentation that notice is being taken of Legacy Mathcad users' requests. Included in the list of such items for inclusion in Mathcad Prime 12 is Plot Titles. Now, I'm all for Plot Titles being added - it is a worthwhile addition - however, I have to question their priority. There is already at least one workaround for plot titles, ie, Text regions (and I bet the new Plot Title feature won't allow embedding of Math regions into titles).
A more worthwhile enhancement priority, in my parochial view, would be the ability to insert a Plot region into a Text Box, in the same way that we can insert Math Regions into Text Boxes.
At the moment, all one can do is paste an image of a plot into a Text Box - having suitably sized it in an external application, such as Paint (not a very satisfactory state of affairs).
The above representation of an embedded plot also shows the utility of being able to hide the define operator ("="). The diagram number (which should have an intrinsic Mathcad autonumber) would look far better without the equals sign.
Stuart
(Oh, in case I forgot to mention it ... expect MEPs for Sumerian/Akkadian/Bablylonian cuneiform, Mayan, and Attic Greek.
Not forgetting 
(I have a book on cuneiform on order. It might help me to determine whether any of the above are combinations of cuneiform symbols. I'm reasonably sure that the last one is the Akkadian locative of water (𒀀𒀀), but my Mesopotamian is a bit rusty).
lsolve is easily implemented in Express (for many, if not most, but certainly not all circumstances) with:
yielding:
Success!
Luc
