Question

这就是我的df的样子，

Region  Dummy value1 value2
Mangonui  NA   NA     NA
Sales     NA   9      6
Kaitaia   NA   NA     NA
Sales     NA   16     1
Whangaroa NA   NA     NA
Sales     NA   2      2

重新生成此步骤的步骤

 structure(list(Region = c("Mangonui", "Sales", "Kaitaia", 
"Sales", "Whangaroa", "Sales"), Dummy = c(NA, 
NA, NA, NA, NA, NA), Dweling_values = c(NA, "9", NA, "16", NA, 
"2"), Section_values = c(NA, "6", NA, "1", NA, "2")), .Names = c("Region", 
"Dummy", "value1", "value2"), row.names = c(NA, 
-6L), class = c("tbl_df", "tbl", "data.frame"))

如何将两行合并在一起，以便我可以获得针对地区名称的销售？所以输出应该是，

Region  Dummy value1 value2
Mangonui  NA   9      6
Kaitaia   NA   16     1
Whangaroa NA   2      2

Answer 1

base R

中的简单重新排列

myNew <- cbind(df$Region[seq.int(1,nrow(df),2)], df[seq.int(2,nrow(df),2), 2:4])
names(myNew) <- names(df)

myNew
     Region Dummy value1 value2
2  Mangonui    NA      9      6
4   Kaitaia    NA     16      1
6 Whangaroa    NA      2      2

的更新
到目前为止，最优雅的解决方案是@thelatemail

cbind(df[1][c(TRUE,FALSE),,drop=FALSE], df[-1][c(FALSE,TRUE),]) Region Dummy value1 value2 1 Mangonui NA 9 6 3 Kaitaia NA 16 1 5 Whangaroa NA 2 2

Answer 2

首先，你需要找到并责骂任何以这种形式提供数据的人。告诉他们，如果他们继续这样做，你就不会成为他们的朋友。然后，只使用一些简单的基本R函数：

# generate indices for the sales and region rows
sales_rows <- seq(2, nrow(df), by = 2)
region_rows <- seq(1, nrow(df), by = 2)

# subset to create the df you really want
sales_df <- df[sales_rows, ]
# use just the names from the region rows
regions <- df[region_rows, "Region"]
sales_df$Region <- regions

# > sales_df
#      Region Dummy value1 value2
# 2  Mangonui    NA      9      6
# 4   Kaitaia    NA     16      1
# 6 Whangaroa    NA      2      2

Answer 3

使用dplyr和tidyr的解决方案。我们的想法是使用recode将Sales替换为NA，使用fill根据之前的行归置NA，然后使用{{1}使用其他列中的任何非NA值过滤行。

filter_at

将下一行合并在一起 - R

3 个答案: