Convert polyline-circle to normal circle
@Anonymous wrote:
i am new to all of this. would love to be able to remove all but 4 vertex's for any and all of my artwork. as for now i need to remove them from a circle. ....
Can you explain further, and/or post a sample drawing with before and after conditions? If you're talking about Circle entities, they don't have vertices, so I'm trying to picture what you really need to do. If you're talking about converting circular Polylines into the same but with fewer vertices, they can have as few as 2 and still be circular -- is there any reason to keep 4 vertices? If you have approximations of circles in something like Polylines with lots of short line segments, those may or may not be easily convertible, depending on how accurately polygonal they are, but see Post 11.
@Anonymous wrote:
i have attached two circles and or letters. that have too many lines, i need a circle or arc, .... it would take me hours to click and remove vertex.
When I bring in your .dxf files, those are already Circles, so you can't remove vertices, and there's nothing to convert. If they look like they're made up of line segments, that's just a matter of view resolution. If you Zoom in and they look segmented, and after a REGEN they look curved, that's all it is.
Type VIEWRES, answer Yes to the question whether you want fast zooms, and then give it a higher number for the circle zoom percent [the default is 100 -- there's not much reason not to give it the maximum of 20000], and you should not have that problem any more unless you Zoom way far out and then back in, but REGEN will always fix their on-screen appearance.
@Anonymous wrote:
they're made up of line segments, very small and i need to remove most of them. ....
That one is a Polyline, at last [none of the previous attachments were, via DXFIN -- they were Circles]. The code in Post 1 will make a Circle from that, but the Polyline is a little taller than it is wide, so the "match" is not perfect. The code in Post 11 will do it, too, but because it does it differently, the "match" is off in a different way. Unlike the Post 1 code, the Post 11 code leaves the Polyline in place, so you can see how they compare, but it could be made to delete the Polyline, as the code in Post 1 does, if you prefer.
@sharpl wrote:
Works! Is there one for ellipse? ....
Are you talking about converting a full Polyline ellipse-approximation drawn with the ELLIPSE command when the PELLIPSE System Variable is set to 1, into a true Ellipse? I can actually imagine a way to do that, because there are characteristics of such Polylines that can be relied on to be regular, from which to define the Ellipse.
But if you're talking about one that's been broken into an elliptical arc [partial ellipse], or, for example, an elliptical shape as a Polyline of loads of short line segments [as a lot of circle-approximations discussed in this thread are], those would be entirely different challenges, if they would be possible at all.
[And by the way -- what "Works!" ? There are a lot of routines on this thread, and even just limiting it to the person you're Replying to, @stevor posted two, in Messages 28 & 30, though maybe they're the same code-wise.]
@Kent1Cooper wrote:
.... Are you talking about converting a full Polyline ellipse-approximation drawn with the ELLIPSE command when the PELLIPSE System Variable is set to 1, into a true Ellipse? I can actually imagine a way to do that....
Just for full ellipses of that kind, without the usual enhancements, and very lightly tested, but here's the way I imagined to do at least that:
;| ELLIPSE-1to0.lsp [command name: E10]
To convert Ellipse(s) from PELLIPSE=1 Polyline approximation to
PELLIPSE=0 true Ellipse.
Works on only full closed "heavy" 2D Polyline ellipses of 16 arc segments, such
as made by ELLIPSE command when PELLIPSE System Variable is set to 1.
Kent Cooper, 8 November 2018
|;
(defun C:E10 (/ ss n el)
(setvar 'pellipse 0)
(if (setq ss (ssget '((0 . "POLYLINE"))))
(repeat (setq n (sslength ss))
(if (= (vlax-curve-getEndParam (setq el (ssname ss (setq n (1- n))))) 16.0)
(progn ; then
(command "_.ellipse"
"_none" (vlax-curve-getStartPoint el)
"_none" (vlax-curve-getPointAtParam el 8)
"_none" (vlax-curve-getPointAtParam el 4)
"_.erase" el ""
); command
); progn
); if
); repeat
); if
(princ)
); defun
For broken ones, it's really hard to think of a way that a routine could find the locations by which to define a true Ellipse. OSNAPping to CENter locations would give different ones for each arc segment involved. It might be possible with enough of them to calculate an effective ellipse center, but I assume you wouldn't always have the same number of arc segments remaining, so that could be very unreliable.
@sharpl wrote:
Broken lines - ellipse in the past- converted to CAD from a .pdf file. I have to redraw the ellipses each time.
Depending on the nature of the conversion from .PDF:
If they're actual partial Ellipse objects, you can use the attached ArcClose.lsp with its AC command to draw the complete Ellipses that they're part of. [Despite the name, it closes partial Ellipses into closed/full ones, as well as closing Arcs into Circles.]
But if they're [for example] Arcs, or arc segments of Polylines, it won't do what you want, but will complete the Circle that they're part of.
[If they're a series of little Line objects somehow, I don't have anything to offer to automate drawing the equivalent Ellipses.]