Hi Eric,<br><br><div class="gmail_quote">On Wed, Apr 14, 2010 at 3:20 PM, Eric E. Monson <span dir="ltr"><<a href="mailto:emonson@cs.duke.edu">emonson@cs.duke.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div style="word-wrap:break-word">Hey Marcus,<div><br></div><div>Tomorrow I'll open bug reports / feature requests on these two recent issues and send you the links, just so you know where they are.</div></div></blockquote>
<div><br></div><div>Sounds great - I will assign them to myself. </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div style="word-wrap:break-word"><div><br></div>
<div>For now, though, I was playing around with the PC chart in ParaView and had another idea that you might want to consider: Right now selections generate an ID list. When you're looking at a time-dependent data set, then, the red selection lines in the PC chart are animated along with the other views when you hit "play", showing how the quantities associated with these IDs change over time. Another useful option would be to keep the range of the original box(es) constant, which would generate a query-type selection, so you could monitor which points or cells exhibit that attribute range. Right now queries can't be chained, which would be necessary to correspond to a multi-axis selection in the PC chart, but I think Utkarsh said this is on the horizon. I don't know how exactly this would work since query-based selections seem to be a ParaView-only thing for now, but I think it would be a nice feature which would fit in well with the parallel coordinates paradigm.</div>
</div></blockquote><div><br></div><div>Yes, I was wondering about the value of such selections. There are a lot of possibilities, and for some of this the vtkSelection infrastructure may need extending. I am hoping to have a chat with Utkarsh when he returns about some of the issues/possibilities with selections in ParaView, along with how far we can go right now. </div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div style="word-wrap:break-word"><div><br></div><div>Oh, and thanks for implementing the PC array selection in the Display tab -- it's _much_ more useful now!</div>
</div></blockquote><div><br></div><div>Yes, this was a very necessary feature. I think that some more API/interface is needed to get things working just right. The opacity spin box can also be very useful for adjusting line opacity/thickness for your data set.</div>
<div><br></div><div>Thanks,</div><div><br></div><div>Marcus </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div style="word-wrap:break-word"><div><div></div><div class="h5">
<div><br><div><div>On Apr 13, 2010, at 9:41 AM, Marcus D. Hanwell wrote:</div><br><blockquote type="cite">Hi Eric,<br><br><div class="gmail_quote">On Fri, Apr 9, 2010 at 3:14 PM, Eric E. Monson <span dir="ltr"><<a href="mailto:emonson@cs.duke.edu" target="_blank">emonson@cs.duke.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
When there are a lot of axes on the parallel coordinates chart, both the axis titles and scale labels run together and are unreadable. Do you think there could be a mode where all labels are suppressed, but you have the hover text like in your line charts? This would let you see both the axis title and value scale without all of the fixed labels cluttering up the chart.<br>
</blockquote><div><br></div><div>This is certainly a nice idea, and is possible to do if I can get some time to implement it. If you open a bug report then I will look into it once if I have time once I am done with higher priority features and bug fixes.</div>
<div><br></div><div>Marcus</div></div></blockquote></div></div></div></div></div></blockquote></div>