R中的自定义断言

Question

我有一个excel工作表数据框架，它包含22列和多行，具体取决于输出。有一列名为(比方说(column_1。现在，我想使用断言测试整个工作表，但有一个条件，如果column_1在它的任何一行(比如第4行(中有一个特定的值1，我编写的断言将不会为所有22列的整行运行，否则它将像往常一样为每列运行断言。

伪数据：

column_1 | column_2 | column_3
NA | "House" | 12
NA | "Plot" | 34
NA | "Office" | 90
1 | "Villa" | 1008
...
...
...

原始逻辑

if (excel_sheet[["column_1"]] == "1") {
# the assertion will **not** run for that specific row
} else {
# the assertion will run for that specific row
assert_numeric(x = excel_sheet[["column_1"]], any.missing = T, lower = 1, upper = 1)
assert_numeric(x = excel_sheet[["column_2"]], any.missing = F, lower = 0, upper = 1)
assert_numeric(x = excel_sheet[["column_3"]], any.missing = F)
assert_character(x = excel_sheet[["column_4"]], any.missing = F)
...
...
...
}

Answer 1

不熟悉checkmate包及其功能。但如果我读对了问题，你想

1. 如果某个特定列没有特定值，请检查该列
2. 如果是这种情况，请运行其余的代码

采用所提供的代码，它应该可以使用for循环：

for (i in excel_sheet$column_1){
if(i != 1){
# comment
assert_numeric(x = i, any.missing = T, lower = 1, upper = 1)
assert_numeric(x = i, any.missing = F, lower = 0, upper = 1)
assert_numeric(x = i, any.missing = F)
assert_character(x = i, any.missing = F)
...
...
...

重要

1. 请注意，注释需要用'#'而不是双斜杠键入
2. 当我阅读你的代码和我的答案时，我觉得代码只关注特定的单元格i，而不是行，你想更改/断言单元格还是行

Answer 2

@Omniswitcher的答案最接近。这是我的最终解决方案。

for (i in 1:nrow(excel_sheet)) {
if (excel_sheet[["column_1"]][i] != 1) {
assert_numeric(x = excel_sheet[["column_2"]][i], any.missing = F, lower = 0, upper = 1)
assert_numeric(x = excel_sheet[["column_3"]][i], any.missing = F)
assert_character(x = excel_sheet[["column_4"]][i], any.missing = F)
...
...
...
}
}