blocks|key|2786949|text|我推荐使用ggplot2。这是一个tutorial。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|2786950|您可能会想要使用grid.arrange()。我经常使用它，它工作得很好。我在上面提供的链接详细介绍了它的用法。|style|CODE|2786951|entityMap|0|LINK|mutability|MUTABLE|url|https://cran.r-project.org/web/packages/egg/vignettes/Ecosystem.html^0|H|8|0|0|8|E|0^^$0|@$1|2|3|4|5|6|7|P|8|@]|9|@$A|Q|B|R|1|S]]|C|$]]|$1|D|3|E|5|6|7|T|8|@$A|U|B|V|F|G]]|9|@]|C|$]]|$1|H|3|-4|5|6|7|W|8|@]|9|@]|C|$]]]|I|$J|$5|K|L|M|C|$N|O]]]]

I would recommend using ggplot2. Here is a <a href="https://cran.r-project.org/web/packages/egg/vignettes/Ecosystem.html" rel="nofollow noreferrer">tutorial</a>.

You will probably want to use <code>grid.arrange()</code>. I have used this frequently and it has worked quite well. The link I provided above gives details on its usage.

blocks|key|70557|text|我很抱歉，如果讨论R中不同绘图机制的优点的旧线程是多余的，那么，我认为可能值得展示具有三种主要绘图机制的“简单”多面板绘图的解决方案。@thelatemail关于将lattice或ggplot2用于这些类型的绘图的建议是正确的，也许这个答案说明了为什么。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|entityRanges|data|70558|尽管基数R可能需要对par()进行最多的调整，并且需要熟悉library(help+=+"graphics")的功能，但我倾向于使用基数R来生成出版物质量数字。在这篇PDF中，我发现了一篇关于Sean+Anderson使用base+R进行多面板绘图的精彩讨论。|70559|首先，在一个data.frame中生成一些可复制的数据(always+a+good+idea)，该数据由18组20个x-y对组成，具有合适的组标签和索引uid。曲线图将显示x-y数据并添加一条平滑线。|BOLD|70560|++set.seed(1234)
++x+<-+seq(0,+9,+length.out+=+20)
++y+<-+replicate(6,+(x-5)+%2B+rnorm(x))
++y+<-+c(y,+replicate(6,+5*sin(x)+%2B+rnorm(x)))
++y+<-+c(y,+replicate(6,+5*atanh((9-x)/10)+%2B+rnorm(x)))

++a+<-+gl(3,+120,+labels+=+c("A","B","C"))+#+these+factors+are+handy
++b+<-+gl(6,+20,+length+=+360)
++uid+<-+as.numeric(b:a)

++df+<-+data.frame(x,+y,+a,+b,+uid)
++rm(x,+y,+a,+b,+uid)+#+prevent+use+of+variables+outside+of+the+data.frame|code-block|syntax|javascript|70561|根据我的经验，R+Studio对情节的要求(控制)更高一些。我不确定这段代码在R+Studio下运行得有多好。出现该警告后，将创建一个大小合适的打印设备以启动。|70562|++dev.new(width+=+6.5,+height+=+6.5)|70563|首先，使用par(mfrow+=+c(6,+3)和外边距参数(oma)的基础R解决方案。这也使得非典型地使用legend()函数来向每个面板添加标题。|70564|++par(mfrow+=+c(6,+3),+mar+=+c(0,0,0,0),+oma+=+c(6,+6,+2,+2))
++ylim+<-+range(df$y)+#+to+ensure+uniformly+sized+plots
++ncol+<-+3+#+number+of+columns+to+the+plot
++for(u+in+1:18)+{
++++sel+<-+df$uid+==+u
++++plot(y+~+x,+df,+subset+=+sel,+ann+=+FALSE,+axes+=+FALSE,+ylim+=+ylim)
++++box()+#+adds+a+"frame"+or+"box"+around+each+plot
++++xy+<-+loess.smooth(df$x[sel],+df$y[sel],+span+=+1/3)
++++lines(xy)
++++if+((u+-+1)%25%25ncol+==+0)+axis(2,+las+=+1)
++++if+((u+-+1)%25/%25ncol+==+5)+axis(1)
++++leg.text+<-+paste(unique(df$a[sel]),+unique(df$b[sel]),+sep+=+":")
++++legend("top",+leg.text,+bty+=+"n")
++}
++mtext("x",+side+=+1,+outer+=+TRUE,+line+=+3)
++mtext("y",+side+=+2,+outer+=+TRUE,+line+=+3)|70565|​|70566|📷|atomic|70567|ggplot2和lattice都将返回可以进一步细化的对象。这里展示了一个基于这种精神的lattice解决方案。|70568|++library(lattice)
++o+<-+xyplot(y+~+x+%7C+b:a,+df,+as.table+=+TRUE,+layout+=+c(3,+6))
++o+<-+update(o,+type+=+c("p",+"smooth"),+span+=+1/3)
++plot(o)|70569|70570|70571|在不同的标绘机制中，语法和嵌入的美学原则是不同的。这在基本的ggplot2解决方案中是显而易见的。|70572|++library(ggplot2)
++g+<-+ggplot(df,+aes(x,+y))+%2B+geom_point()+%2B+facet_wrap(b:a+~+.,+ncol+=+3)
++g+<-+g+%2B+geom_smooth()
++plot(g)|70573|70574|70575|选择使用哪种机制是个人的选择。我提到过，为了更好的控制，我通常更喜欢基础图形。为了准备要在监视器上显示的绘图，ggplot2提供了屏幕友好的图像，几乎不需要大惊小怪。对于重复探索多维数据，我发现lattice是最有用的。这可以通过执行绘图代码(不加载库)所需的时间来说明。这里(从Windows+i7机器)捕获并显示了每个文件的system.time()输出。|70576|++rbind(base+=+time1,+lattice+=+time2,+ggplot+=+time3)[,+1:3]
>+++++++++user.self+sys.self+elapsed++
>+base+++++++++0.05+++++0.03++++0.10++
>+lattice++++++0.23+++++0.03++++0.28++
>+ggplot+++++++1.27+++++0.05++++1.38++|70577|entityMap|0|LINK|mutability|MUTABLE|url|https://seananderson.ca/courses/11-multipanel/multipanel.pdf|1|https://seananderson.ca/|2|IMAGE|IMMUTABLE|imageUrl|https://ask.qcloudimg.com/http-save/yehe-900000/ed40a2c9729ff51e23fbc2c0798affa6.png|imageAlt|3|https://ask.qcloudimg.com/http-save/yehe-900000/01cdff46a330f72614f7c48665ec64ef.png|4|https://ask.qcloudimg.com/http-save/yehe-900000/c393675e9660cad1c5a5a8662789c09f.png^0|2A|7|2I|7|0|A|5|T|Q|2B|3|0|2O|D|1|0|S|6|24|3|0|0|0|0|5|J|V|3|1I|8|0|0|0|0|1|2|0|0|7|8|7|18|7|0|0|0|0|1|3|0|U|7|0|0|0|0|1|4|0|1J|7|2P|7|4K|D|0|0^^$0|@$1|2|3|4|5|6|7|22|8|@$9|23|A|24|B|C]|$9|25|A|26|B|C]]|D|@]|E|$]]|$1|F|3|G|5|6|7|27|8|@$9|28|A|29|B|C]|$9|2A|A|2B|B|C]]|D|@$9|2C|A|2D|1|2E]|$9|2F|A|2G|1|2H]]|E|$]]|$1|H|3|I|5|6|7|2I|8|@$9|2J|A|2K|B|J]|$9|2L|A|2M|B|C]]|D|@]|E|$]]|$1|K|3|L|5|M|7|2N|8|@]|D|@]|E|$N|O]]|$1|P|3|Q|5|6|7|2O|8|@]|D|@]|E|$]]|$1|R|3|S|5|M|7|2P|8|@]|D|@]|E|$N|O]]|$1|T|3|U|5|6|7|2Q|8|@$9|2R|A|2S|B|C]|$9|2T|A|2U|B|C]|$9|2V|A|2W|B|C]]|D|@]|E|$]]|$1|V|3|W|5|M|7|2X|8|@]|D|@]|E|$N|O]]|$1|X|3|Y|5|6|7|2Y|8|@]|D|@]|E|$]]|$1|Z|3|10|5|11|7|2Z|8|@]|D|@$9|30|A|31|1|32]]|E|$]]|$1|12|3|13|5|6|7|33|8|@$9|34|A|35|B|C]|$9|36|A|37|B|C]|$9|38|A|39|B|C]]|D|@]|E|$]]|$1|14|3|15|5|M|7|3A|8|@]|D|@]|E|$N|O]]|$1|16|3|Y|5|6|7|3B|8|@]|D|@]|E|$]]|$1|17|3|10|5|11|7|3C|8|@]|D|@$9|3D|A|3E|1|3F]]|E|$]]|$1|18|3|19|5|6|7|3G|8|@$9|3H|A|3I|B|C]]|D|@]|E|$]]|$1|1A|3|1B|5|M|7|3J|8|@]|D|@]|E|$N|O]]|$1|1C|3|Y|5|6|7|3K|8|@]|D|@]|E|$]]|$1|1D|3|10|5|11|7|3L|8|@]|D|@$9|3M|A|3N|1|3O]]|E|$]]|$1|1E|3|1F|5|6|7|3P|8|@$9|3Q|A|3R|B|C]|$9|3S|A|3T|B|C]|$9|3U|A|3V|B|C]]|D|@]|E|$]]|$1|1G|3|1H|5|M|7|3W|8|@]|D|@]|E|$N|O]]|$1|1I|3|-4|5|6|7|3X|8|@]|D|@]|E|$]]]|1J|$1K|$5|1L|1M|1N|E|$1O|1P]]|1Q|$5|1L|1M|1N|E|$1O|1R]]|1S|$5|1T|1M|1U|E|$1V|1W|1X|-4]]|1Y|$5|1T|1M|1U|E|$1V|1Z|1X|-4]]|20|$5|1T|1M|1U|E|$1V|21|1X|-4]]]]

I apologize if this is redundant with older threads discussing the merits of the different plotting mechanisms in R. However, I think it may be worth showing solutions for "simple" multipanel plots with the three major plotting mechanisms. The recommendation of @thelatemail about using <code>lattice</code> or <code>ggplot2</code> for these type of plots is right on the money and perhaps this answer illustrates why.

Still, even though base R may require the most tweaking through <code>par()</code> and requires familiarity with the functions found with <code>library(help = "graphics")</code>, I tend to use base R to generate publication quality figures. I found an excellent discussion of multipanel plotting with base R by <a href="https://seananderson.ca/" rel="nofollow noreferrer">Sean Anderson</a> in this <a href="https://seananderson.ca/courses/11-multipanel/multipanel.pdf" rel="nofollow noreferrer">PDF</a>.

To start, generate some reproducible data (always a good idea) composed of 18 sets of 20 x-y pairs in a single data.frame with suitable group labels and an index <code>uid</code>. The plots will show the x-y data and add a smoothing line.

<pre><code> set.seed(1234)
 x &lt;- seq(0, 9, length.out = 20)
 y &lt;- replicate(6, (x-5) + rnorm(x))
 y &lt;- c(y, replicate(6, 5*sin(x) + rnorm(x)))
 y &lt;- c(y, replicate(6, 5*atanh((9-x)/10) + rnorm(x)))

 a &lt;- gl(3, 120, labels = c("A","B","C")) # these factors are handy
 b &lt;- gl(6, 20, length = 360)
 uid &lt;- as.numeric(b:a)

 df &lt;- data.frame(x, y, a, b, uid)
 rm(x, y, a, b, uid) # prevent use of variables outside of the data.frame
</code></pre>

In my experience, R Studio is bit more demanding (controlling) with regards to plots. I'm not sure how well this code will run under R Studio. With that warning, an appropriately sized plotting device will be created to start.

<pre><code> dev.new(width = 6.5, height = 6.5)
</code></pre>

First, a base R solution using <code>par(mfrow = c(6, 3)</code> and the outer margin parameter (<code>oma</code>). This also makes atypical use of the <code>legend()</code> function to add a titles to each panel.

<pre><code> par(mfrow = c(6, 3), mar = c(0,0,0,0), oma = c(6, 6, 2, 2))
 ylim &lt;- range(df$y) # to ensure uniformly sized plots
 ncol &lt;- 3 # number of columns to the plot
 for(u in 1:18) {
 sel &lt;- df$uid == u
 plot(y ~ x, df, subset = sel, ann = FALSE, axes = FALSE, ylim = ylim)
 box() # adds a "frame" or "box" around each plot
 xy &lt;- loess.smooth(df$x[sel], df$y[sel], span = 1/3)
 lines(xy)
 if ((u - 1)%%ncol == 0) axis(2, las = 1)
 if ((u - 1)%/%ncol == 5) axis(1)
 leg.text &lt;- paste(unique(df$a[sel]), unique(df$b[sel]), sep = ":")
 legend("top", leg.text, bty = "n")
 }
 mtext("x", side = 1, outer = TRUE, line = 3)
 mtext("y", side = 2, outer = TRUE, line = 3)
</code></pre>

<a href="https://i.stack.imgur.com/1X0KK.png" rel="nofollow noreferrer"><img src="https://i.stack.imgur.com/1X0KK.png" alt="base multipanel plot"></a> 
Both <code>ggplot2</code> and <code>lattice</code> will return objects that can be further refined. A solution with <code>lattice</code> in that spirit is shown here.

<pre><code> library(lattice)
 o &lt;- xyplot(y ~ x | b:a, df, as.table = TRUE, layout = c(3, 6))
 o &lt;- update(o, type = c("p", "smooth"), span = 1/3)
 plot(o)
</code></pre>

<a href="https://i.stack.imgur.com/LJzLU.png" rel="nofollow noreferrer"><img src="https://i.stack.imgur.com/LJzLU.png" alt="lattice multipanel plot"></a> 
The syntax as well as the embedded aesthetic principles differ among the plotting mechanisms. This is obvious in a basic <code>ggplot2</code> solution.

<pre><code> library(ggplot2)
 g &lt;- ggplot(df, aes(x, y)) + geom_point() + facet_wrap(b:a ~ ., ncol = 3)
 g &lt;- g + geom_smooth()
 plot(g)
</code></pre>

<a href="https://i.stack.imgur.com/yXppD.png" rel="nofollow noreferrer"><img src="https://i.stack.imgur.com/yXppD.png" alt="ggplot2 multipanel plot"></a> 
The choice of which mechanism to use is personal one. I mentioned that I often favor base graphics for fine control. For preparing plots to be displayed on a monitor, <code>ggplot2</code> gives screen-friendly images with little fuss. For repeatedly exploring multidimensional data, I find <code>lattice</code> to be the most useful. This can be illustrated by the time required to execute the plotting code (without loading the library). The output of <code>system.time()</code> for each was captured and shown here (from a Windows i7 machine).

<pre><code> rbind(base = time1, lattice = time2, ggplot = time3)[, 1:3]
&gt; user.self sys.self elapsed 
&gt; base 0.05 0.03 0.10 
&gt; lattice 0.23 0.03 0.28 
&gt; ggplot 1.27 0.05 1.38 
</code></pre>

I am making about 18 graphs in R and I want them to render in one image. I get an error message about size. 

In the past I have used <code>par(mfrow = c(2,2))</code> and such simple stuff to generate multiple graphs in one rendering. 

So, in what I am doing I want to generate a six times three rendering. So, it would look something like six rows and three columns. I thought it should be simple, by using <code>par(mfrow = c(6,3))</code>.

In R when I try to do <code>par(mfrow = c(6,3))</code> I get 

<blockquote>
 "Error in plot.new() : figure margins too large" .
</blockquote>

I tried to solve this problem by using 3 times 3 instead of 6 times 3. This would result in half of what I want, but then I can just combine them through MSpaint. A bit of a hassle, but it does the job to produce the final 6 times 3. Is there an easier way to do this, to allow this to be done in R?

generating graphs in one image

Windows

我在R中制作了大约18个图形，我想让它们在一个图像中渲染。我收到一条关于大小的错误消息。在过去，我使用par(mfrow = c(2,2))和这种简单的东西在一次渲染中生成多个图形。因此，在我所做的工作中，我希望生成一个6乘以3的渲染。因此，它看起来像六行三列。我认为这应该很简单，通过使用par(mfrow = c(6...

问在一个图像中生成图形
EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问在一个图像中生成图形EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问在一个图像中生成图形
EN