r - 计算因素的出现并在 R 中的行和列中对它们求和

Question

我有这张表:

smoke <- matrix(c("one"," ","three","four "," "," ",
              " ","two","three","four"," ","Other",
              "one"," ","three","four"," "," ",
              " "," "," "," ","none"," ",
              " "," "," "," "," "," ",
              "  "," ","three","four"," ","Other",
              " "," "," "," ","none"," "
 ),ncol=6,byrow=TRUE)
colnames(smoke) <- c("Var_1","Var_2", "Var_3","Var_4","None","Other")
rownames(smoke) <- c(1,2,3,4," ",6,7)

烟:

      Var_1  Var_2   Var_3    Var_4   None   Other  

  1    one           three     four 
  2           two    three     four          Other
  3    one           three     four
  4                                   None

  6                  three     four          Other
  7                                   None

现在我想:

1. 删除空行(例如第 5 行)。数据集的行太多，我可以轻松检测到空行。
2. 计算行和列中出现的次数。您可能已经注意到，每列中的字符串(变量)是相同的。即相同的名称，相同的长度。但是，“无”列中的值在列总和中应视为 0。

这是我想要的结果:

        Var_1   Var_2    Var_3     Var_4     None    Other   col_sum

 1      one              three     four                          3
 2              two      three     four              Other       4
 3      one              three     four                          3
 4                                           None                0
 5                       three     four              Other       3
 6                                           None                0
 rowSum  2       1         4        4         2         2

如果这些是数字数据，我已经学会了如何处理类似的问题。但是现在，我不确定如何解决它。感谢您的帮助!

Answer 1

我们可以在矩阵上使用trimws，在逻辑矩阵上用rowSums计算非空

# // trim out the leading/lagging spaces
smoke <- trimws(smoke)
# // subset the rows by removing rows having only blank
smoke1 <- smoke[!!rowSums(smoke != ""),]
# // get the count of non-blank, non-none elements
rowSums(smoke1 != "" & smoke1 != "none")
#1 2 3 4 6 7 
#3 4 3 0 3 0
#// and with columns, use `colSums`
colSums(smoke1 != "")
#  Var_1 Var_2 Var_3 Var_4  None Other 
#    2     1     4     4     2     2