More explicit usage of explicit -function

[TikkaMasala]

Hi,

I would like to present clearly all intermediate steps of computations to my students.

In the attached sheet I try to explain the problem that I'm having with the explicit-function.

Is it possible to show intermediate results within the computations using the explicit -function, or should I use some other funcition?

I'm having a lecture tomorrow, so I would appreciate a quick reply

[MikeArmstrong]

If you are only trying to show your students the computations steps of the calculation and not trying to teach them the Mathcad symbolic functions, I don't see any reason for using symbolic's.

Mike

[TikkaMasala]

Hi Mike,

Thank you for reply.

I don't try to teach them mathcad, but geotechnics, so mathcad is only a tool to show computations more clearly.

I totally agree with you when a computation is as simple as the attachement was,

but when a computation get's more complicated and when formulas get longer,

it is really hard to follow only variables and try to remember what each of them stands for.

Showing intermediate steps could help the situation, or at least if I would do the computations by hand,

I would show intermediate steps as well.

-matti

[MikeArmstrong]

Just an idea - I would lay it out as below.

This way the students can directly see which variable relates to which value.

Mike

[TikkaMasala]

If I make the computation sheet using vectors, is it somehow possible to update the values in the second row just by changin the input values,

such as gamma (unit weight of soil) in my example?

[MikeArmstrong]

If I make the computation sheet using vectors, is it somehow possible to update the values in the second row just by changin the input values,

such as gamma (unit weight of soil) in my example?

unfortunately not, well, not that I know of.

Will you be using SI standard units throughout? Because if so there might be another way.

Mike

[TikkaMasala]

Yes, I'm using only SI -units.

[MikeArmstrong]

I had this very same problem when completing an Open University course. The coursework had to be submitted showing workings, unfortunately I had to do the work long hand by manually pasting the results into the second row of the vector.

I'm sure this could be solved by a bit of programming, but I totally forgot about writing something until now.

I will try and get a sheet working, but in the interim period you might have to do it long hand unless another forum member can help.

Sorry.

Mike

[TikkaMasala]

Thank you for your effort.

I hope you can remember how this programming bit was done at some point, since I have still many lectures to teach and this could be something that many of my colleagues would appreciate as well

[PhilipOakley]

It took a bit of finding, but I think it is possible.

It is two thirds the way down the Explicit Calculation Help

C:\Program Files\Mathcad\Mathcad 14\Resource Center\en\qsheet\tutorial\features\explicit.xmcd

You need to use the symbolic calculation at the appropriate stage so that you get a redefinition so the value isn't traced back to the source.

I'm getting some 'red' after the redefinition which may be because of greek characters .. ... no doesn't look like it's that...

see sheet.

Philip

[RichardJ]

If you want it to simplify the units, then no, it will not do that. The explicit keyword is handled by the symbolic processor, and the symbolic processor is not units aware. It just sees kN and m as undefined variables.

[MikeArmstrong]

I would still like Mathcad to be able to strip units from a variables and keep track of them. Similar to the SIUnitsOf(x) function. Maybe ExtractUnitOf(x) and then the function could convert the unit into a string or hold them as a constant.

I'm not quite sure what I would want the function to do, but an improvement on extracting units would be nice.

Mike

[PhilipOakley]

The point is that the symbolic answer (an expression) should be assigned to the variable (including the units; Pa and m), which can then be evaluated by the numeric processor, which does know the numeric multipliers of the units.

The new numeric answer should then be available for the next step in the explicit substitution chain. I have logged a call with the fault, as this should have a sensible solution, and the tutorials say it can be done!

The main thing is to get PTC to log the problem and then to resolve it in the next 6 monthly update (they did promise a regular release schedule didn't they?)

Philip

[RichardJ]

The main thing is to get PTC to log the problem and then to resolve it in the next 6 monthly update (they did promise a regular release schedule didn't they?)

I don't think that is a simple fix. At the moment the symbolic results can be passed to the numeric processor within a single expression. I think that to add another step so it goes symbolic processor, numeric processor, symbolic processor, and somehow the explicit keyword (which is inherently a symbolic operation, not a numeric one) still works, would be far from trivial.

[PhilipOakley]

The fundamental problem is, as Richard has mentioned, with units. But it's slightly more complicated I think.

The issue is that when we chain the assignment of an explicit evaluation with the numeric evaluation, we get an assignment of the final (right hand) result which has been reduced to its simplest unit - the pascal Pa in this case.

Unfortunately the assignment doesn't properly recognise the perfectly valid expression, and thinks that the Pa is undefined, even though if you type it in manually, both the numeric and symbolic processors are happy with it.

Even if you predefine it (Pa) yourself, it isn't seen, so it suggest that the internal representation (from the numeric result) isn't being recognised (by the symbolic processor) rather than the the displayed string.

I've supplied a sheet to PTC support.

Philip

PS a side point: when you use an assignment on an explicit multiply, you need to include the assigned variable name in the explicit, not just teh multiplied variables.

b:=2 c:=3

a := b.c explict, b, c -> a Doesn't work as expected

a := b.c explicit, a, b, c -> 2.3 Does work.