Hi. 2D Vector Field plot

vfield.sm (109 КиБ) скачан 234 раз(а).
vfield.pdf (193 КиБ) скачан 242 раз(а).

Best regards.
Alvaro.

Sorry Alvaro: all red SS6179

You can use a supplement that makes the lines monochrome.

vfieldMonochrome.sm (55 КиБ) скачан 155 раз(а).

Wrote

Sorry Alvaro: all red SS6179

Hi Jean. I use range instead for loops. This feature is available only in newer versions, don't know starting from which one.
Best regards.
Alvaro.

Wrote

You can use a supplement that makes the lines monochrome.

Hi Ber. Thanks very much for the tip. You cycle all colors to return to the first one. Amazing creativity!

Best regards.
Alvaro.

Improved with the Ber suggestion. Also, some explanation about the vectors added.

vfield.sm (124 КиБ) скачан 140 раз(а).

Best regards.
Alvaro.

Wrote

Improved with the Ber7 suggestion. Also, some explanation about the vectors added.

Thanks Alvaro, good job. SS6179 complains for the last two modules.
Added Tom code if it can help for more difficult example(s).

Cheers ... Jean

ODE rkf Integrate [rkf4 2 parameters Algo] Else.sm (205 КиБ) скачан 122 раз(а).

Wrote

Added Tom code if it can help for more difficult example(s).

Hi Jean. Thanks for the examples. This is my version of the Tom's Example.



vfield Tom Gutman's IntegralCurves Example.sm (137 КиБ) скачан 145 раз(а).

Best regards.
Alvaro.

Thank you Alvaro, impressive work.

Jean's example



Best regards.
Alvaro.

Wrote

Hi Jean. Thanks for the examples. This is my version of the Tom's Example.

Really impressive ! Thanks for your great enhancement of SS.
SS 6179 complains, but easy to get the integral path otherwise,
though for this example it is short of going past the "Wolf line".

Jean



vfield Tom Gutman's IntegralCurves Example.sm (144 КиБ) скачан 125 раз(а).

Wrote

Really impressive ! Thanks for your great enhancement of SS.

In some way(s), I reconciliated myself with your vector field.
I understand the multiple "ic" feature, that I couldn't doctor.
Your VF is so far superior to MCD native "Vector Fields",
though: there are lot of specifications that don't present well.

Cheers ... Jean



VectorField [Alvaro]_2.sm (31 КиБ) скачан 132 раз(а).

Numerical Contour integral, as it figure in the mathcad quick sheets.

Numerical Contour Integral - SMath.pdf (85 КиБ) скачан 127 раз(а).
Numerical Contour Integral - SMath.sm (57 КиБ) скачан 142 раз(а).
Numerical Contour Integral - Mathcad.pdf (91 КиБ) скачан 119 раз(а).

Best regards.
Alvaro.

It takes a lot, but it looks great

ElectricField.sm (127 КиБ) скачан 143 раз(а).



Best regards.
Alvaro.

An update employing Viacheslav's suggestion from here to force them to be monochrome traces, although it only works in the XYPlot plugin and not in SMath 2D charts. I imagine that he must know it very well because in reality it is not that the plugin is his, but also that the procedure for the implicit graph is taken from the mathcad forum from a post by him.



ElectricField XYPltot.sm (132 КиБ) скачан 114 раз(а).

I'm still thinking that even that it could be great to have NaN (Not A Number) as an another choice.

Best regards.
Alvaro.

Wrote

I'm still thinking that even that it could be great to have NaN (Not A Number) as an another choice.

Tom has used 10^307 as NaN doing same as Mathcad built-NaN. Not tested Smath.

Wrote

Wrote

I'm still thinking that even that it could be great to have NaN (Not A Number) as an another choice.

Tom has used 10^307 as NaN doing same as Mathcad built-NaN. Not tested Smath.

Hi Jean. Thanks for the tip. I know it. Also Tom Gutman use NaN = i and others more complicated like int(1,NaN), this is, NaN as the differential. But that before the update for Mathcad 11 2a. This version was hard to crack, and students have the previous version 11 2b, without NaN, or next releases, where mathcad starts it's dead, but all of them with NaN. Now 11 2a release it's dead because it's hardware authentication involving the net card and how know what's else.

In SMath NaN = big number is useless in plots because don't break the line, and NaN = i returns an error in plots. Without implementing a complicated programming structure for NaN, taking an empty string "" as break when plotting nx2 matrices was enough, I think, for avoiding the need to cycle colors for keep the same.

XY plot can emulate that, but you need to change the data structure, as matrices inside matrices. That's make complicate further use of the data points.

Best regards.
Alvaro.

X-Y Plot uses NaN internally in order to hide points that can't be drawn. You can ask Andrey to add the ability to enter this "number". We have symbolic infinity, why not have a NaN?

Hola Alvaro,
All what you said about Mathcad is right. Yes, Tom used various versions of NaN.
My Mathcad 8 Pro was a crack before spending $ ... Worked perfect.
Consequently from beta testing 2001i, Mathsoft send me the 2001i CD.
Then for developing DATA Pack, Mathcad 11. Few days after 11, came
a several MB update in two modules. There was 2 versions of 11.2a.
Enterprise Edition and Single User Edition.
Enterprise Edition unprotected by Macrovision.
Single User Edition, protected by Macrovision.
On May 30 2015 15:30, PTC ruptured their contract with Macrovision,
all Single User Edition got zapped. Impossible to crack Macrovision,
as you said, there is no crack for MCD 11.2a.
Though, I got zapped MCD 11.2a, I still have it running on XP Pro
from a wise back to back installation by my Technician before the killing date.
Possibly, Macrovision is simply a hidden times stamp ?
Cheers ... Jean

Wrote

X-Y Plot uses NaN internally in order to hide points that can't be drawn. You can ask Andrey to add the ability to enter this "number". We have symbolic infinity, why not have a NaN?

Hello Viacheslav . Thanks for the recommendation, and sorry for the delay. I do not make that recommendation because I am not sure what the priorities may be. Implementing NaN can be quite complicated, and it should have an effect on almost all plugins, as they should be able to accept that new "number". NaN is regulated by too rigid standards, and its implementation may not be urgent. If it were up to me, I could fill out a very long list of requests, and I am sure that Andrey would attend to many of them, as he has done before, but I am aware that some of the things that I may ask for would be impractical, as kind.

I could ask that instead of implementing NaN I take empty strings, or empty string arrays as a special kind of NaN, but I don't know how much programming effort something that would be a temporary solution might cost.

In short, I do not order because I do not know the priorities. Now, if you ask me to risk an opinion, not a request, I think that what should be prioritized is to have a robust and predictable numerical linear algebra library.

Best regards.
Alvaro.