Re: [R] Contour plot on a triangular mesh
On 01/20/2012 11:28 PM, Roary wrote: Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Hi Roary, Have a look at the triax.fill function in the plotrix package. Jim __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Contour plot on a triangular mesh
Thanks for your responses Jason and Jim. The ternary plot is definitely the right style with the triangle fill. However, one of the important features (that perhaps I understated) is that X3 will have more categories than X1 and X2, therefore the triangular shape is not appropriate. If X1 has I categories and X2 has J categories, then X3 has I+J-1 categories. Therefore, the overall plot would be in the shape of a parralelogram. Almost like an extended ternary plot... The goal is to view these unbalanced categories on the same scaling as each other... Would this be possible to construct? Thanks again, Roary. Jim Lemon wrote On 01/20/2012 11:28 PM, Roary wrote: Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Hi Roary, Have a look at the triax.fill function in the plotrix package. Jim __ R-help@ mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://r.789695.n4.nabble.com/Contour-plot-on-a-triangular-mesh-tp4313098p4315772.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Contour plot on a triangular mesh
On 01/21/2012 09:38 PM, Roary wrote: Thanks for your responses Jason and Jim. The ternary plot is definitely the right style with the triangle fill. However, one of the important features (that perhaps I understated) is that X3 will have more categories than X1 and X2, therefore the triangular shape is not appropriate. If X1 has I categories and X2 has J categories, then X3 has I+J-1 categories. Therefore, the overall plot would be in the shape of a parralelogram. Almost like an extended ternary plot... The goal is to view these unbalanced categories on the same scaling as each other... Would this be possible to construct? Hi Roary, Possible, yes. Easy, no, unfortunately. The attached PDF shows unequal categories drawn using the triax.abline function. It's easy enough to get the edges by simply drawing lines across the plot. Filling the resulting polygons would take quite a bit of calculation to work out the segments necessary to draw each one. The equilateral triangles in triax.fill are just drawn at equal spacings, edges across the x axis and heights down the y axis. Alternatively, triax.plot and triax.fill could be rewritten to use a non-equilateral triangle as the plotting field, which would have internal triangles of the same proportions. Either way it means a fair amount of calculation, which I can't do at the moment. If you would like to work out the trigonometry, I could probably knock it over... Jim triax_test.pdf Description: Adobe PDF document __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Contour plot on a triangular mesh
Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Many thanks, Roary -- View this message in context: http://r.789695.n4.nabble.com/Contour-plot-on-a-triangular-mesh-tp4313098p4313098.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Contour plot on a triangular mesh
Not sure if I understand the question: If you have more data the grid produced by image() or contour() will be finer anyway... Perhaps we just need an example what you are actually asking for. Uwe Ligges On 20.01.2012 13:28, Roary wrote: Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Many thanks, Roary -- View this message in context: http://r.789695.n4.nabble.com/Contour-plot-on-a-triangular-mesh-tp4313098p4313098.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Contour plot on a triangular mesh
I have two observed categorical variables X1 and X2, with X3=X1+X2, and a continuous response Y. I can interpolate the surface and construct an ordinary 2D square contour plot (with X1,X2 axes and X3 on the diagonal). However, I would like to change the orientation of the plot so that the axes fit a parallelogram shaped grid made up from triangles. This would place X3 on the same scale as X1 and X2 and allow for an easier interpretation of the data for my research question. I hope this makes more sense! Thanks, Roary. Uwe Ligges-3 wrote Not sure if I understand the question: If you have more data the grid produced by image() or contour() will be finer anyway... Perhaps we just need an example what you are actually asking for. Uwe Ligges On 20.01.2012 13:28, Roary wrote: Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Many thanks, Roary -- View this message in context: http://r.789695.n4.nabble.com/Contour-plot-on-a-triangular-mesh-tp4313098p4313098.html Sent from the R help mailing list archive at Nabble.com. __ R-help@ mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@ mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://r.789695.n4.nabble.com/Contour-plot-on-a-triangular-mesh-tp4313098p4314337.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Contour plot on a triangular mesh
I'm not sure if this is appropriate. If the sum of your variables is always the same constant, then you might try a Ternary plot ( http://en.wikipedia.org/wiki/Ternary_plot ). The vcd package can make ternary plots. On 01/20/2012 02:56 PM, Roary wrote: I have two observed categorical variables X1 and X2, with X3=X1+X2, and a continuous response Y. I can interpolate the surface and construct an ordinary 2D square contour plot (with X1,X2 axes and X3 on the diagonal). However, I would like to change the orientation of the plot so that the axes fit a parallelogram shaped grid made up from triangles. This would place X3 on the same scale as X1 and X2 and allow for an easier interpretation of the data for my research question. I hope this makes more sense! Thanks, Roary. Uwe Ligges-3 wrote Not sure if I understand the question: If you have more data the grid produced by image() or contour() will be finer anyway... Perhaps we just need an example what you are actually asking for. Uwe Ligges On 20.01.2012 13:28, Roary wrote: Hi All, I have 3 variables which present a perfect linear dependency such that the third is the sum of the first two. I have an ordinary 2D contour plot on a square grid with the first two variables forming the axes and the third naturally being the diagonals. From an interpretive point of view it would be nice to plot these two variables on a finer grid such that the third can have the same scaling (i.e. a finer grid) as the first two and this would look better on a triangular mesh. Is this possible in R? Many thanks, Roary __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.