<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">Hello Federico,<br>
<br>
Le 06/04/2019 à 00:27, Federico Miyara a écrit :<br>
</div>
<blockquote
cite="mid:5e6e75ad-7228-4106-9345-0d1a28d0a612@fceia.unr.edu.ar"
type="cite">
<meta http-equiv="content-type" content="text/html;
charset=windows-1252">
<br>
Dear All,<br>
<br>
<font face="Courier New">I have the following problem with the
graphic axes property clip_state. I don't know if comes from an
inaccurate description in the documentation or something I'm not
understanding quite well...<br>
<br>
1) I create new axes by plotting some data set. As expected,
axes' size matches the plot data range. <br>
<br>
2) I run <br>
<br>
gca().clip_state = "off"<br>
<br>
Acording to the documentation (<a moz-do-not-send="true"
class="moz-txt-link-freetext"
href="https://help.scilab.org/docs/6.0.2/en_US/axes_properties.html">https://help.scilab.org/docs/6.0.2/en_US/axes_properties.html</a>),
from now on all newly created objects should not clip, meaning
that even if the object (such as a polyline) exceeds the
original data range, it will be displayed in full. <br>
<br>
3) I plot another data set that exceeds the original axes
limits. Contrary to the expected behavior, it does clip.<br>
<br>
4) I run <br>
<br>
gca().clip_state <br>
<br>
and I see it has reverted to "clipgrf". This explains the
observed clipping, but it is not clear why the setting from 2)
has changed.<br>
<br>
5) I run again<br>
<br>
gca().clip_state = "off"<br>
<br>
Nothing happens (it makes sense, since the clip_state should
affect new objects, not previous ones)<br>
<br>
6) I run <br>
<br>
gce().children.clip_state = "off"<br>
<br>
Only now the clipping is gone and the second plot shows in full.
<br>
<br>
Why the original clip_state setting is overridden? <br>
<br>
If it is the normal or expected behavior (in which case the
documentation isn't accurate), then what is it meant for, given
that the final impact is that of each individual entity's
setting?<br>
</font></blockquote>
<br>
Confirmed.<br>
<br>
This behavior occurs for ages: It was already the case with Scilab
4.0.<br>
It occurs even with <tt>gca().auto_scale = "off"</tt>.<br>
It comes from plot2d(), that resets the value to "clipgrf".<br>
It's not the case with xstring() or xpoly(), or i guess with other
"low level" graphic functions like xarcs() etc.<br>
<br>
I don't know if such a reset is intentional, and if so, why.<br>
Anyway, it is not (yet) reported as a bug on Bugzilla.<br>
<br>
Regards<br>
Samuel<br>
<font face="Courier New"><br>
</font>
</body>
</html>