hello again.

How can I get a graph from a variable? Example in graph 2 where nothing appears.

the content of the variable is pasted directly into graph 1 and
there it is working

I generate the matrices to draw a series of rectangles
but I don't know in advance how many rectangles will be drawn
that's why I require to use a variable that contains a system of several matrices

is there another solution?

stack-of-rec.sm (29,67 KiB) scaricato 1232 volta(e).

There might be other issues, but one of them could be the bug demonstrated here. Normally, no element assignment should alter the object type from list to matrix as it seems to happen in the line block definition. I think this is another regression, such stuff used to work in older versions (here 7251).

The original desparate workaround with str2num tries to handle the bug. Normally you should be able to reliably initialize a variable using any object of appropriate type.


Lists.sm (7,5 KiB) scaricato 1215 volta(e).

Wrote

is there another solution?

Transit the 'gg' matrices via file, plot.

stack-of-rec.sm (31,93 KiB) scaricato 1210 volta(e).

thank you very much Martin Kraska and also to you Jean Giraud

both options work well
I attach the file with Martin's suggestion, which I
It seems what I was looking for, and I take note of the solution
Jean who as always is very helpful and educational
since you learn a lot from these examples.

Thank you!

stack-of-rec.sm (23,12 KiB) scaricato 1201 volta(e).

Much simpler Amigo ...
Isolate 'gg' gg:= gg ...or rectangles:= gg

Wrote

Much simpler Amigo ...
Isolate 'gg' gg:= gg ...or rectangles:= gg

Hahaha yes! this work too.

If you want unicolor plot and rotation,
you may have to reconsider your construction.

Utilities XY [Draw Truss].sm (22,85 KiB) scaricato 1043 volta(e).
stack-of-rec.sm (37,11 KiB) scaricato 1212 volta(e).

Your H beam, you can have it SVG style.

Spline De Casteljau SVG Hbeam.sm (41,36 KiB) scaricato 1297 volta(e).

Wrote

There might be other issues, but one of them could be the bug demonstrated here. Normally, no element assignment should alter the object type from list to matrix as it seems to happen in the line block definition. I think this is another regression, such stuff used to work in older versions (here 7251).

The original desparate workaround with str2num tries to handle the bug. Normally you should be able to reliably initialize a variable using any object of appropriate type.

The issue demonstrated in the above post nails down to unexpected handling of single element lists used to initialize the data under numeric optimization.



lists2.sm (6,44 KiB) scaricato 1200 volta(e).