cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Help us improve the PTC Community by taking this short Community Survey! X

Symbolic Solve Fails to use solve variable as argument

Strib
11-Garnet

Symbolic Solve Fails to use solve variable as argument

Mathcad Prime 7 symbolic solve block fails to use a solved variable as argument to built-in function.

 

It's common to need solve for one or more variables and then express some function of one or more variables.  E.g., two angles are solved symbolically then the sine of the sum of the two angles is sought.

 

MC Prime gives silly result with, apparently, any built-in function.  This is the case whether a solve variable is equated to a symbolic or a constant.

 

I attach an illustration using the simplest thing I can imagine:  a solve variable is equated to something (a symbolic value or a constant), then a variable is equated to a function of the first solve variable.

 

I offer four simple solves, each solving for an explicitly written function of the stated value and also a solve for the function of the first solve variable as the argument. 

 

Is this inherent in MC Prime - that it can't use solve variables in a function?

 

 

ACCEPTED SOLUTION

Accepted Solutions
Strib
11-Garnet
(To:Werner_E)

Thanks Werner_E for validating me.  I was afraid it's a bug in Prime 7 symbolics.  I didn't know you had a choice in Prime 6 of one that works and one that doesn't.

 

My illustration was brainless trivial... but you can imagine the pitiful mess with a a complicated set of symbolic equations.

 

I'll be happy to notify PTC - how do I do that?

View solution in original post

6 REPLIES 6
Werner_E
25-Diamond I
(To:Strib)

I guess you found yet another bug in Primes "new" symbolics.

In Prime 6 we have the choice between the older legacy symbolic (muPad) and the new one (an Axiom fork).

Here is what the lecgacy symbolic gives:

Werner_E_0-1662062389567.png

 

The result of the new symbolic simply is wrong! After all, all you are doing (or trying to do) is to solve a simple system of equations with three  equations in three variables.

Werner_E_1-1662062475693.png

If you feel like doing so and think that it would help, you may consider reporting this bug to PTC.

 

The attached pdf shows what your sheet looks like when evaluated using the legacy symbolic (not available anymore starting with Prime 7).

 

Observation: If you change the order of the variables after "solve" you get a correct result even with the new symbolics:

Werner_E_1-1662063871784.png

Strib
11-Garnet
(To:Werner_E)

Thanks Werner_E for validating me.  I was afraid it's a bug in Prime 7 symbolics.  I didn't know you had a choice in Prime 6 of one that works and one that doesn't.

 

My illustration was brainless trivial... but you can imagine the pitiful mess with a a complicated set of symbolic equations.

 

I'll be happy to notify PTC - how do I do that?

Werner_E
25-Diamond I
(To:Strib)


I'll be happy to notify PTC - how do I do that?

 


You'll have to open a support case /support ticket  whatever they call it. https://support.ptc.com/apps/case_logger_viewer/cs/auth/ssl/log

As PTC is not interested in bug reports by non-paying customers, you will only be able to do so if you are still  under maintenance (which should be the case anyway given that PTC now only sells time limited lease licenses and no perpetual licenses anymore).

 

EDIT: As far as I am aware of, some bugs in the symbolics were fixed in Version P8. I don't have P8 installed so I cannot tell if this bug still exists in P8. Maybe someone with a full P8 installation can tell us more.

Here is an announcement of the improvements I found:
Improvements to the symbolic engine

Just like the numeric engine, Mathcad Prime 8 brings you new symbolic engine functionality, including:

  • Further enhancements to symbolic integral transforms, with new use cases and better performance, such as:
    • Fourier and invfourier
    • Laplace and invlaplace
    • Ztrans and invztrans
  • Define normalisation and oscillatory factors with new nfact and ofact modifiers that you use with fourier keywork
  • Enhanced performance and new supported use cases for calculus operators (limits, derivative, integrals)
  • More use cases and improved performance for keywords (solve, rewrite, series, simplify, assume)
  • Undefined variable handling functionality improved in symbolic results, including better automated labelling of free undefined variables and lambda expressions

 

DJNewman
17-Peridot
(To:Werner_E)

On Prime 8, using the last example as an example, I get the silly result:

 

DJNewman_0-1662065969129.png

 

Reordering seems to resolve it:

DJNewman_1-1662066086662.png

If I were you, I would submit this as a bug and see what happens.

I manage the Creo and PTC Mathcad YouTube channels for PTC, as well as all PTC Mathcad marketing in general.
Strib
11-Garnet
(To:DJNewman)

Thanks, DJNewman.  Sad to see Prime 8 has the same quirk.

With a lengthy set of equations to relate it wouldn't always be possible to order the solve variables as to which ones do 'outputs' and which ones need 'inputs'.

 

I'll endeavor to submit a bug report.  But at the moment I'm locked out - I'm a legal purchaser back in February but PTC rejects my Product Code.  I just posted a help request regarding this.

Werner_E
25-Diamond I
(To:Strib)

You shouldn't post the product code in a public forum. I'd suggest that you edit your post and delete that code (or ask a moderator to do so for you if you are not allowed to edit your post anymore).

Announcements

Top Tags