The function merge_column_names() repairs typically Microsoft Excel file imports where, in the raw data, column headers are spread over multiple rows and/or include merged cells.
merge_column_names(df, name_rows, first_data_row = NULL, sep = " ", drop_name_rows = TRUE)
df must be a data frame
name_rows should be an integer vector (or coercable to one) which specifies the row numbers in which all desired column header information exists. These do not need to be presented in numerical order but instead in the order in which you want the contents of each row concatenated in a single column header. Currently they must be contiguous but a future update will allow them to be non-contiguous.
first_data_row defaults to the row below the largest integer in name_rows but can be overridden.
sep defines what should separate each individual value in combined header texts.
drop_name_rows by default removes all rows named in named_rows and also all rows up to but not including first_data_row.
Values in merged cells are allocated to all cells to the right of the initial value where rows which include separate columns grouped under a merged cell are specified within name_rows. Empty cells are also filled rightwards with the value of the preceding (i.e. leftwards) non-empty cell in the same row. Any value within name_rows which contains a numerical digit at the end of a word is assumed to be a footnote character and is removed.
Returns a data frame.