python 用于绘制三元/三角形图的库/工具
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/701429/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Library/tool for drawing ternary/triangle plots
提问by Cristian Ciupitu
I need to draw ternary/triangle plotsrepresenting mole fractions (x, y, z) of various substances/mixtures (x+ y+ z= 1). Each plot represents iso-valued substances, e.g. substances which have the same melting point. The plots need to be drawn on the same triangle with different colors/symbols and it would be nice if I could also connect the dots.
我需要绘制表示各种物质/混合物 ( x+ y+ z= 1) 的摩尔分数 ( x, y, z) 的三元/三角形图。每幅图代表等值物质，例如具有相同熔点的物质。这些图需要用不同的颜色/符号绘制在同一个三角形上，如果我也可以连接点就好了。
I have looked at matplotlib, R and gnuplot, but they don't seem to be able to draw this kind of plot. The 3rd party ade4package for R seems to be able to draw it, but I'm not sure if I can draw multiple plots on the same triangle.
我看过 matplotlib、R 和 gnuplot，但他们似乎无法绘制这种情节。R的第 3 方ade4包似乎能够绘制它，但我不确定是否可以在同一个三角形上绘制多个图。
I need something that runs under Linux or Windows. I'm open to any suggestions, including libraries for other languages, e.g. Perl, PHP, Ruby, C# and Java.
我需要在 Linux 或 Windows 下运行的东西。我愿意接受任何建议，包括其他语言的库，例如 Perl、PHP、Ruby、C# 和 Java。
采纳答案by doug
R has an external package called VCDwhich should do what you want.
R 有一个名为VCD的外部包，它应该可以满足您的需求。
The documentation is very good (122 page manual distributed w/ the package); there's also a book by the same name, Visual Display of Quantitative Information, by the package's author (Prof. Michael Friendly).
文档非常好（122 页手册随包分发）；还有一本同名的书，定量信息的可视化显示，由包的作者（Michael Friendly 教授）编写。
To create ternary plots using vcd, just call ternaryplot()and pass in an m x 3 matrix, i.e., a matrix with three columns.
要使用vcd创建三元图，只需调用ternaryplot()并传入一个 mx 3 矩阵，即具有三列的矩阵。
The method signature is very simple; only a single parameter (the m x 3 data matrix) is required; and all of the keyword parameters relate to the plot's aesthetics, except for scale, which when set to 1, normalizes the data column-wise.
方法签名非常简单；只需要一个参数（mx 3 数据矩阵）；并且所有关键字参数都与绘图的美感有关，但 scale 除外，当设置为 1 时，会按列对数据进行标准化。
To plot data points on the ternary plot, the coordinates for a given point are calculated as the gravity center of mass pointsin which each feature value comprising the data matrix is a separate weight, hence the coordinates of a point V(a, b, c) are
为了在三元图上绘制数据点，给定点的坐标被计算为质点的重心，其中包含数据矩阵的每个特征值是一个单独的权重，因此点的坐标 V(a, b, c) 是
V(b, c/2, c * (3^.5)/2
To generate the diagram below, i just created some fake data to represent four different chemical mixtures, each comprised of varying fractions of three substances (x, y, z). I scaled the input (so x + y + z = 1) but the function will do it for you if you pass in a value for its 'scale' parameter (in fact, the default is 1, which i believe is what your question requires). I used different colors & symbols to represent the four data points, but you can also just use a single color/symbol and label each point (via the 'id' argument).
为了生成下图，我刚刚创建了一些假数据来表示四种不同的化学混合物，每种化学混合物由三种物质（x、y、z）的不同比例组成。我缩放了输入（所以 x + y + z = 1）但是如果你为它的 'scale' 参数传入一个值（事实上，默认值是 1，我相信这就是你的问题）需要）。我使用不同的颜色和符号来表示四个数据点，但您也可以只使用一种颜色/符号并标记每个点（通过“id”参数）。
回答by dave
Created a very basic script for generating ternary (or more) plots. No gridlines or ticklines, but those wouldn't be too hard to add using the vectors in the "basis" array.
创建了一个非常基本的脚本来生成三元（或更多）图。没有网格线或刻度线，但使用“基础”数组中的向量添加这些不会太难。
from pylab import *
def ternaryPlot(
data,
# Scale data for ternary plot (i.e. a + b + c = 1)
scaling=True,
# Direction of first vertex.
start_angle=90,
# Orient labels perpendicular to vertices.
rotate_labels=True,
# Labels for vertices.
labels=('one','two','three'),
# Can accomodate more than 3 dimensions if desired.
sides=3,
# Offset for label from vertex (percent of distance from origin).
label_offset=0.10,
# Any matplotlib keyword args for plots.
edge_args={'color':'black','linewidth':2},
# Any matplotlib keyword args for figures.
fig_args = {'figsize':(8,8),'facecolor':'white','edgecolor':'white'},
):
'''
This will create a basic "ternary" plot (or quaternary, etc.)
'''
basis = array(
[
[
cos(2*_*pi/sides + start_angle*pi/180),
sin(2*_*pi/sides + start_angle*pi/180)
]
for _ in range(sides)
]
)
# If data is Nxsides, newdata is Nx2.
if scaling:
# Scales data for you.
newdata = dot((data.T / data.sum(-1)).T,basis)
else:
# Assumes data already sums to 1.
newdata = dot(data,basis)
fig = figure(**fig_args)
ax = fig.add_subplot(111)
for i,l in enumerate(labels):
if i >= sides:
break
x = basis[i,0]
y = basis[i,1]
if rotate_labels:
angle = 180*arctan(y/x)/pi + 90
if angle > 90 and angle <= 270:
angle = mod(angle + 180,360)
else:
angle = 0
ax.text(
x*(1 + label_offset),
y*(1 + label_offset),
l,
horizontalalignment='center',
verticalalignment='center',
rotation=angle
)
# Clear normal matplotlib axes graphics.
ax.set_xticks(())
ax.set_yticks(())
ax.set_frame_on(False)
# Plot border
ax.plot(
[basis[_,0] for _ in range(sides) + [0,]],
[basis[_,1] for _ in range(sides) + [0,]],
**edge_args
)
return newdata,ax
if __name__ == '__main__':
k = 0.5
s = 1000
data = vstack((
array([k,0,0]) + rand(s,3),
array([0,k,0]) + rand(s,3),
array([0,0,k]) + rand(s,3)
))
color = array([[1,0,0]]*s + [[0,1,0]]*s + [[0,0,1]]*s)
newdata,ax = ternaryPlot(data)
ax.scatter(
newdata[:,0],
newdata[:,1],
s=2,
alpha=0.5,
color=color
)
show()
回答by Nicholas Hamilton
A package I have authored in R has justbeen accepted for CRAN, webpage is www.ggtern.com:
我在 R 中编写的一个包刚刚被 CRAN 接受，网页是www.ggtern.com：
It is based off ggplot2, which I have used as a platform. The driving force for me, was a desire to have consistency in my work, and, since I use ggplot2 heavily, development of the package was a logical progression.
它基于ggplot2，我将其用作平台。我的驱动力是希望在我的工作中保持一致性，并且由于我大量使用 ggplot2，因此包的开发是合乎逻辑的。
For those of you who use ggplot2, use of ggtern should be a breeze, and, here is a couple of demonstrations of what can be achieved.
对于那些使用 ggplot2 的人来说，使用 ggtern 应该是轻而易举的，这里有几个可以实现的演示。
Produced with the following code:
使用以下代码生成：
# Load data
data(Feldspar)
# Sort it by decreasing pressure
# (so small grobs sit on top of large grobs
Feldspar <- Feldspar[with(Feldspar, order(-P.Gpa)), ]
# Build and Render the Plot
ggtern(data = Feldspar, aes(x = An, y = Ab, z = Or)) +
#the layer
geom_point(aes(fill = T.C,
size = P.Gpa,
shape = Feldspar)) +
#scales
scale_shape_manual(values = c(21, 24)) +
scale_size_continuous(range = c(2.5, 7.5)) +
scale_fill_gradient(low = "green", high = "red") +
#theme tweaks
theme_tern_bw() +
theme(legend.position = c(0, 1),
legend.justification = c(0, 1),
legend.box.just = "left") +
#tweak guides
guides(shape= guide_legend(order =1,
override.aes=list(size=5)),
size = guide_legend(order =2),
fill = guide_colourbar(order=3)) +
#labels and title
labs(size = "Pressure/GPa",
fill = "Temperature/C") +
ggtitle("Feldspar - Elkins and Grove 1990")
Contour plots have also been patched for the ternary environment, and, an inclusion of a new geometry for representing confidence intervals via the Mahalanobis Distance.
还为三元环境修补了等高线图，并且包含了用于通过Mahalanobis 距离表示置信区间的新几何图形。
Produced with the following code:
使用以下代码生成：
ggtern(data=Feldspar,aes(An,Ab,Or)) +
geom_confidence(aes(group=Feldspar,
fill=..level..,
alpha=1-..level..),
n=2000,
breaks=c(0.01,0.02,0.03,0.04,
seq(0.05,0.95,by=0.1),
0.99,0.995,0.9995),
color=NA,linetype=1) +
geom_density2d(aes(color=..level..)) +
geom_point(fill="white",aes(shape=Feldspar),size=5) +
theme_tern_bw() +
theme_tern_nogrid() +
theme(ternary.options=element_ternary(padding=0.2),
legend.position=c(0,1),
legend.justification=c(0,1),
legend.box.just="left") +
labs(color="Density",fill="Confidence",
title="Feldspar - Elkins and Grove 1990 + Confidence Levels + Density") +
scale_color_gradient(low="gray",high="magenta") +
scale_fill_gradient2(low="red",mid="orange",high="green",
midpoint=0.8) +
scale_shape_manual(values=c(21,24)) +
guides(shape= guide_legend(order =1,
override.aes=list(size=5)),
size = guide_legend(order =2),
fill = guide_colourbar(order=3),
color= guide_colourbar(order=4),
alpha= "none")
回答by xioxox
Veuszsupports ternary plots. Here is an example from the documentation:
回答by Frédéric Grosshans
Chlo? Lewis developed a triangle-plot general class, meant to support the soil texture trianglewith Python and Matplotlib. It's available here http://nature.berkeley.edu/~chlewis/Sourcecode.htmlhttps://github.com/chlewissoil/TernaryPlotPy
克洛？Lewis 开发了一个三角图通用类，旨在用 Python 和 Matplotlib支持土壤质地三角形。它可以在这里找到http://nature.berkeley.edu/~chlewis/Sourcecode.html https://github.com/chlewissoil/TernaryPlotPy
Chloe editing to add: Moved it to a more reliable host! Also, it's a public repo, so if you want to request library-ization, you could add an issue. Hope it's useful to someone.
Chloe 编辑添加：将其移至更可靠的主机！此外，它是一个公共存储库，因此如果您想请求图书馆化，您可以添加一个问题。希望它对某人有用。
回答by Phil H
回答by Deaton
I just discovered a tool which uses Python/Matplotlib to generate ternary plots called wxTernary. It's available via http://wxternary.sourceforge.net/-- I was able to successfully generate a ternary plot on the first try.
我刚刚发现了一个使用 Python/Matplotlib 生成名为 wxTernary 的三元图的工具。它可以通过http://wxternary.sourceforge.net/ 获得——我第一次尝试就成功地生成了一个三元图。
回答by Eric
There is a R package named soiltexture. It's aimed at soil texture triangle plot, but can be customized for some aspects.
有一个名为soiltexture的 R 包。它针对土壤质地三角形图，但可以针对某些方面进行定制。
回答by CookieOfFortune
Find a vector drawing library and draw it from scratch if you can't find an easier way to do it.
如果找不到更简单的方法，请查找矢量绘图库并从头开始绘制。
