J
Jerry W. Lewis
I need to regularly import data for statistical analysis from documents
containing multiple Word tables that have numbered footnotes. A single
footnote may apply to multiple table entries which makes it difficult to
distinguish the footnote number from additional digits in the reported data.
I do not control data or table creation process, but may be able to recommend
simple changes.
Currently, I Paste Special|Text the Word document into Excel and then read
the Excel worksheet with SAS. Pasting into Excel produces a much more
consistent layout for import, since it effectively unmerges merged cells in
the table. As a bonus, pasting also drops the footnotes and original
footnote numbers. The problem is that it does not drop footnote numbers from
footnote cross-references, regardless of whether they are explicitly entered
as footnote cross-references, or are simply supers-scripted numbers that
merely look like footnote cross-references.
SAS imports via the MS supplied Jet engine, which cannot distinguish
formatting, so the superscripting, while useful for human readers seems
totally useless to the import process.
Any suggestions for modifying either the Word document creation process or
the data import process to minimize the manual labor of removing stray
footnote numbers from data values and verifying that no footnote numbers have
contaminated the data set?
Jerry
containing multiple Word tables that have numbered footnotes. A single
footnote may apply to multiple table entries which makes it difficult to
distinguish the footnote number from additional digits in the reported data.
I do not control data or table creation process, but may be able to recommend
simple changes.
Currently, I Paste Special|Text the Word document into Excel and then read
the Excel worksheet with SAS. Pasting into Excel produces a much more
consistent layout for import, since it effectively unmerges merged cells in
the table. As a bonus, pasting also drops the footnotes and original
footnote numbers. The problem is that it does not drop footnote numbers from
footnote cross-references, regardless of whether they are explicitly entered
as footnote cross-references, or are simply supers-scripted numbers that
merely look like footnote cross-references.
SAS imports via the MS supplied Jet engine, which cannot distinguish
formatting, so the superscripting, while useful for human readers seems
totally useless to the import process.
Any suggestions for modifying either the Word document creation process or
the data import process to minimize the manual labor of removing stray
footnote numbers from data values and verifying that no footnote numbers have
contaminated the data set?
Jerry