具有多个从属变量的表

Question

如何使用以下数据创建表格：

初始数据表：

结束表结果：

这里的代码很容易输入R（我会在运行期间从csv文件中读取）：

Date = c(20170101,20170101,20170102,20170102,20170102,20170102,20170103)
Person = c("Adam","Adam","Adam","Ben","Ben","Ben","Ben")
Fruit = c("Apple", "Pear","Bananna","Blueberry","Cherry","Grape","Grape")
z= data.frame(Date,Person,Fruit)

Answer 1

tapply()函数可以使用c函数将值排列到矩阵对象中。值被封装在列表中：

with(z, tapply(Fruit, list(Person,Date), FUN=c))

     20170101    20170102    20170103
Adam Character,2 "Bananna"   NULL    
Ben  NULL        Character,3 "Grape" 

然后，您可以使用pander显示该矩阵对象：

library(pander)
panderOptions('keep.line.breaks', TRUE)
mytable <- with(z, tapply(Fruit, list(Person,Date), FUN=c))
pandoc.table(mytable, style="multiline")

----------------------------------------------------------
  &nbsp;    20170101           20170102          20170103 
---------- ----------- ------------------------ ----------
 **Adam**  Apple, Pear         Bananna             NULL   

 **Ben**      NULL     Blueberry, Cherry, Grape   Grape   
----------------------------------------------------------

如果你想要分隔“单元格”的线条，你可以使用style =“grid”。然后输出如下：

+----------+-------------+--------------------------+------------+
|  Person  |  20170101   |         20170102         |  20170103  |
+==========+=============+==========================+============+
|   Adam   | Apple, Pear |         Bananna          |            |
+----------+-------------+--------------------------+------------+
|   Ben    |             | Blueberry, Cherry, Grape |   Grape    |
+----------+-------------+--------------------------+------------+

如果您将paste0与折叠一起使用，则可以删除逗号并改为使用换行符：

mytable <- with(z, tapply(Fruit, list(Person,Date), FUN=paste0,  collapse="\n"))

矩阵的结构现在是不同的，每个条目都是长度为1的字符值，但是然后它会显示在一个pandoc表中，每个Fruit都在一行上：

pandoc.table(mytable, style="multiline")  # or style ="grid" as above
#------------
-------------------------------------------
       20170101   20170102   20170103 
---------- ---------- ---------- ----------
 **Adam**    Apple     Bananna       NA    
              Pear                         

 **Ben**       NA     Blueberry    Grape   
                        Cherry             
                        Grape              
-------------------------------------------