1-AUG-2019 SolveSys - Equation Library and Nonlinear Systems Solver

[min21]

Hello,
When I hit the plot, there is an error saying "Check exactly one equation."
I installed the equation library app through merged version.
HP connectivity toolkit Build: 2016 12.8 Rev:11226
Thanks.

[Han]

(03-02-2017 10:00 PM)min21 Wrote:  Hello,
When I hit the plot, there is an error saying "Check exactly one equation."
I installed the equation library app through merged version.
HP connectivity toolkit Build: 2016 12.8 Rev:11226
Thanks.

That error message is never used in the app. That error messages is reported by the built-in app. So seeing that message suggests that the app was not properly installed. Can you give some details as to how you installed the merged version? Go into the app, press [Shift] [1] to see the Program Browser menu. At the very top should be an entry that ends with (App) -- it should be called Equation Library (App) if you installed using the hpappdir format. Open it and it should show a version number toward the top.

[min21]

(03-02-2017 11:36 PM)Han Wrote:  That error message is never used in the app. That error messages is reported by the built-in app. So seeing that message suggests that the app was not properly installed. Can you give some details as to how you installed the merged version? Go into the app, press [Shift] [1] to see the Program Browser menu. At the very top should be an entry that ends with (App) -- it should be called Equation Library (App) if you installed using the hpappdir format. Open it and it should show a version number toward the top.

Hello, Han.
I just updated my calculator again using the toolkit. And reinstalled the Equation library. It worked this time. Thank you for your help and for this amazing app.

[akmon]

A work of art this reference manual, could be part of the HP Prime´s users´ guide.

[EdDereDdE]

Where can I get the HP Prime Keys font please?

[Han]

(03-06-2017 04:57 PM)EdDereDdE Wrote:  Where can I get the HP Prime Keys font please?

http://www.hpmuseum.org/forum/thread-1925.html

[EdDereDdE]

Thanks Han

[compsystems]

IDEA: For the creation of the equation library, we can contact the author of formlua PRO to generate a common database..

forum FormulaPro
https://tiplanet.org/forum/viewtopic.php?f=50&t=9576

Formula ProGenerator
https://tiplanet.org/scripts/FormulaProGenerator/

[Han]

(03-06-2017 05:45 PM)compsystems Wrote:  IDEA: For the creation of the equation library, we can contact the author of formlua PRO to generate a common database..

forum FormulaPro
https://tiplanet.org/forum/viewtopic.php?f=50&t=9576

Formula ProGenerator
https://tiplanet.org/scripts/FormulaProGenerator/

I think that it would be great to have formulas and accompanying documentation of the formulas. To that end, users are more than welcome to submit their own *.lib files and I will be happy to maintain them (i.e. update them so that any format changes will be applied to the library files) and include them with the app. Just putting together the list of equations in from the original Equation Library was a tedious task, and I suspect that there might still be typos yet to be discovered. It is not something I want to spend much time on. (I ended up having to learn about black body radiation, fluid dynamics, and a number of other topics just to ensure that the equations from the original Equation Library were fully compatible in the Prime version. Learning is fun, but I prefer choosing my own topics rather being forced into any particular one.) Any future changes to the format of the *lib files will have an accompanying "update" utility/program so that anyone can manage their own library as well. However, that is the extent of my interest in terms of maintenance of the equations themselves.

Right now, all the equations and related data are stored in "Equation Library.lib" -- the next update will allow users to have multiple *.lib files (e.g. one for Chemistry only, another for Physics only, etc.) to help organize their formulas. The next update will also feature the images that were used in the old Equation Library on the HP48 series in addition to allowing users to "switch" libraries. Those images are already in the most recent reference manual (see first post).

[math7]

Thanks to Han for this important program !!! I was testing some equations and in the electricity section when I try to solve resistors in parallel, enter the values for r1 and r2 and when I try to solve Rp and Rs, the program shows me the message "Error evaluating F!".

[math7]

Thanks to Han for this important program !!! I was testing some equations and in the electricity section when I try to solve resistors in parallel, enter the values for r1 and r2 and when I try to solve Rp and Rs, the program shows me the message "Error evaluating F!".

[Han]

(03-09-2017 04:56 AM)math7 Wrote:  Thanks to Han for this important program !!! I was testing some equations and in the electricity section when I try to solve resistors in parallel, enter the values for r1 and r2 and when I try to solve Rp and Rs, the program shows me the message "Error evaluating F!".

\( \mathbf{F} \) here refers to the vector
\[ \mathbf{F} = \begin{bmatrix} f_1(x_1, x_2, \dotsm, x_m) \\ f_2(x_1, x_2, \dotsm, x_m) \\
\vdots \\ f_n(x_1, x_2, \dotsm, x_m) \\ \end{bmatrix} \]

where each \( f_i(x_1,x_2,\dotsm , x_m) \) corresponds to the \(i\)-th equation \( f_i(x_1,x_2, \dotsm, x_m) = 0 \). In other words, if your equation was \( x_1+x_2 = 5 \) then \( f(x_1,x_2) = x_1+x_2 - 5 \) is the function equivalent. So what does the message mean when there is an error evaluating \( \mathbf{F} \)? Look at your initial guess for \( R_p \) -- it's 0. This results in division by 0 in the equation involving \( R_p \). This just means you picked a poor initial guess -- try a value that does not lead to division by 0.

[math7]

Yes it is correct Han, when giving other values to Rs and Rp there is no problem, the results are correct !!! Thank you very much...

[akmon]

Fantastic upgrade. This app is now a must have for Hp Prime. As Eric Reching says in his web: "Get it".

As a suggestion, a shorcut to the picture will improve the information. I remember to use it often even inserting values, it helps a lot. So if there were a quick way to watch the picture (for example, long press on the screen, or a virtual button on the screen or physical button) both in the equation menu and variable menu will be useful.

[StephenG1CMZ]

Certainly a great app.
Just for information: Microsoft's PDF Reader on Windows 8.1 renders manual.pdf as 63 blank pages (or else rendering is very slow)....
Perhaps it would be useful to keep the installation instructions separate.

[Han]

(03-13-2017 10:16 PM)StephenG1CMZ Wrote:  Certainly a great app.
Just for information: Microsoft's PDF Reader on Windows 8.1 renders manual.pdf as 63 blank pages (or else rendering is very slow)....
Perhaps it would be useful to keep the installation instructions separate.

There should have been an install.txt (or did I name it readme.txt?) file separate from the manual. I will make sure that it is included in the next release.

[StephenG1CMZ]

(03-13-2017 10:57 PM)Han Wrote:  

(03-13-2017 10:16 PM)StephenG1CMZ Wrote:  Certainly a great app.
Just for information: Microsoft's PDF Reader on Windows 8.1 renders manual.pdf as 63 blank pages (or else rendering is very slow)....
Perhaps it would be useful to keep the installation instructions separate.

There should have been an install.txt (or did I name it readme.txt?) file separate from the manual. I will make sure that it is included in the next release.

Sorry, yes there is a readme- I should have read it first.

[Han]

KNOWN ISSUES (v0.900)

1. Deleting systems fails when there are more than 70 systems of equations. This is due to a limitation of the input() command, which only allows up to 7 rows per page, and 10 pages max. The current equation library has over 100 systems. There will be a workaround in an upcoming update.

2. Not really an issue, but the library file can be made slightly smaller by removing image width and height parameters (can use grobh and grobw to find the heights and widths). There will be a slight modification to the *.lib file format, and animations will be supported in the next update.

[Han]

So it appears the HP Prime's built-in Solve app has the same issue as the HP48's multiple equation solver (MES): they do not properly analyze inconsistent systems, and return "zeros" that are actually not solutions to the system. Here's an example for the HP48:

http://www.hpmuseum.org/forum/thread-776...l#pid69187

The same problem occurs on the HP Prime. Ask it to solve X+Y=5 and X+Y=6. Set Y to 3 and solve for X to get a "zero" of X=2 and Y=3. I think the problem could be easily fixed by having the Solve app actually check to see if it is indeed a zero of the system, and not just one of the equations. My plan was to include a the Solve app's solver engine as option (among three: quasi-Newton engine in SolveSys, Solve app solver, and a hybrid of the two). It's hard to include it as an option when it fails to check if the system is inconsistent.

[jorge_rui]

I've registered,
Just to say thank you!