按组计算真/假值的条形图（闪避图） [英] Barchart of count of true/false values by group (dodged graphs)

问题描述

我在R和ggplot中很新。我不确定我想要的是否可行。

以下是我的一部分数据：

 > MDF 
批次ABCDE 
 1个G FALSE TRUE FALSE TRUE FALSE 
 2 G FALSE FALSE FALSE TRUE FALSE 
的3G FALSE TRUE FALSE FALSE FALSE 
 4个G FALSE FALSE FALSE TRUE FALSE 
 5个G FALSE FALSE TRUE TRUE TRUE 
 6G的FALSE FALSE TRUE TRUE TRUE 
 7个G FALSE FALSE FALSE FALSE TRUE 
 8个G的FALSE FALSE TRUE TRUE TRUE 
 9分配ģFALSE FALSE FALSE TRUE FALSE 
 10个G FALSE FALSE FALSE TRUE TRUE 
图11G FALSE FALSE FALSE FALSE TRUE 
将12g FALSE FALSE FALSE TRUE FALSE 
将13g FALSE FALSE FALSE TRUE FALSE 
地下14 FALSE FALSE FALSE TRUE FALSE 
 15政FALSE FALSE FALSE TRUE FALSE 
地下16 FALSE FALSE FALSE TRUE FALSE 
 17克FALSE FALSE TRUE TRUE FALSE 
 18G的FALSE FALSE TRUE TRUE TRUE 
 19 A FALSE FALSE FALSE TRUE TRUE 
 20 A FALSE FALSE FALSE TRUE TRUE 
  

其中Batch可以是A，B，G，R， S和其他列（AE）都是布尔/逻辑值。

我可以创建一个图表，它可以计算列B中的TRUE值，如下所示：

解决方案

是的，您可能需要先重新整理数据，然后才能使用 position =dodge为每个键绘制一个栏。 tidyr ：

  library（tidyr）
 library（ dplyr）
 library（ggplot2）
 
 mdf％>％gather（key，value，-Batch）％>％
 ggplot（。，（aes（Batch，as .nu​​meric（value），fill = key）））+ 
 stat_summary（fun.y = sum，geom =bar，position =dodge）
  pre> 
I am rather new in R and ggplot. And I am not sure if what I want is doable.

Here is (a portion of) my data:
> mdf
   Batch     A     B     C     D     E
1      G FALSE  TRUE FALSE  TRUE FALSE
2      G FALSE FALSE FALSE  TRUE FALSE
3      G FALSE  TRUE FALSE FALSE FALSE
4      G FALSE FALSE FALSE  TRUE FALSE
5      G FALSE FALSE  TRUE  TRUE  TRUE
6      G FALSE FALSE  TRUE  TRUE  TRUE
7      G FALSE FALSE FALSE FALSE  TRUE
8      G FALSE FALSE  TRUE  TRUE  TRUE
9      G FALSE FALSE FALSE  TRUE FALSE
10     G FALSE FALSE FALSE  TRUE  TRUE
11     G FALSE FALSE FALSE FALSE  TRUE
12     G FALSE FALSE FALSE  TRUE FALSE
13     G FALSE FALSE FALSE  TRUE FALSE
14     G FALSE FALSE FALSE  TRUE FALSE
15     G FALSE FALSE FALSE  TRUE FALSE
16     G FALSE FALSE FALSE  TRUE FALSE
17     G FALSE FALSE  TRUE  TRUE FALSE
18     G FALSE FALSE  TRUE  TRUE  TRUE
19     A FALSE FALSE FALSE  TRUE  TRUE
20     A FALSE FALSE FALSE  TRUE  TRUE
where Batch can be any of A, B, G, R, S, and the other columns (A-E) are all boolean/logical values.

I was able to create a graph which counts the TRUE values in column B as follows:


using:
ggplot(data = mdf, aes(x = Batch, y = as.numeric(B), fill = Batch))
    + stat_summary(fun.y = sum, geom = "bar")
Similarly, I can easily create 4 more graphs for the other columns (A, C, D, E).

But, is it possible to 'merge' these 5 graphs in a single graph? In other words, I would like a graph where I would have 5 groups (as in the above graph) for the 5 values of 'Batch', and in each group I would need a separate bar with the count of each of the 5 columns (A-E). Is this doable?

Update: Here is what I was looking for (created after using the suggestion from @mtoto)
解决方案
Yes it's possible, you just need to reshape your data first, then you can use position = "dodge" to draw a bar for each key. With tidyr:
library(tidyr)
library(dplyr)
library(ggplot2)

mdf %>% gather(key, value, -Batch) %>%
  ggplot(.,(aes(Batch, as.numeric(value), fill = key))) +
  stat_summary(fun.y = sum, geom = "bar", position = "dodge")


                        
这篇关于按组计算真/假值的条形图（闪避图）的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！