Powerful convenience for Julia visualizations and data analysis http://juliaplots.github.io/
[slack] <chen.tianc> maybe not latex
[slack] <asinghvi17> Yeah you can show unicode no sweat…we use DejaVu by default which supports a lot of this
[slack] <chen.tianc> All I want to do, is to mouse click the screen, and add species, and gradually build large chemical reaction network graphically
[slack] <asinghvi17> If you click on the screen, then you can obtain a position which you can do arbitrary things with
[slack] <chen.tianc> ok, I see. That’s great
[slack] <asinghvi17> yep…Simon did
[slack] <mschauer> do you have a picture for me?
[slack] <asinghvi17> https://simondanisch.github.io/ReferenceImages/gallery/streamplot_3d/index.html
[slack] <loki._.> aka, how to best fix JuliaPlots/Plots.jl#1859
[slack] <loki._.> 👍
[slack] <mkborregaard> I get this
[slack] <mkborregaard> Do you mean why is the origin not exactly at 0?
[slack] <mkborregaard> That’s because the default in Plots is to give a little space so the lines don’t overlap the axis. I believe that’s the default in most languages, though notably not in matplotlib
[slack] <mkborregaard> pass
ylim if you want another behaviour
[slack] <ericphanson> I think he means that the lines don’t extend all the way to (0,0), not the position of the axes
[slack] <ericphanson> which is different than what you would get if you passed a 0:0.01:10 or something for the a x points
[slack] <mkborregaard> ah I see
[slack] <mkborregaard> yes, I think there’s an issue for that
[slack] <mkborregaard> would be good to fix
[slack] <yakir12> In regular plots, when I plot a line described by a bunch of xy coordinates, interpolating a straight line between those coordinates makes sense. A straight line between two subsequent coordinates is a real possibility. In a polar plot I feel this is not the case. A line between two subsequent coordinates can't be interpolated as a straight line because the coordinate system is not a Cartesian map, it is polar. It should be an arc instead of a straight line.
[slack] <yakir12> The straight lines in this plot are wrong in that exact sense. They "should" be arcs. Don't gte me wrong, I don't mean that the package I used to plot this is wrong somehow (in this case it was PGFPlotsX.jl). But I'm trying to raise this as a more general issue, in polar plots in general.
[slack] <simon> One could also argue for the cartesian case, that just connecting dots with a straight line, "is the wrong interpolation"
[slack] <mkborregaard> Plots hides the backend interface. it’s either-or
[slack] <mkborregaard> but if you have a specific keyword that you think is missing and that can easily be shunted through to at least pyplot, gr and one other, it’s very easy to write a PR adding that keyword
[slack] <loki._.> hmm.. my target is a keyword, that only makes sense for tex-output
[slack] <yakir12> Yes, there could be a situation where a linear interpolation is wrong for a Cartesian map. My argument though is that "the right thing to do" for a Cartesian map is most commonly a linear interpolation. The same interpolation however is most commonly wrong for a polar map. So while yes, I guess there could be cases where a linear interpolation is wrong for Cartesian and right for polar, those cases are equally rare. And thus it is downright silly to plot straight lines between discrete points in a polar plot, just as it is silly to assume that a user will want arcs between their Cartesian coordinates.