topic How to read PDF File in SAS EG in SAS Enterprise Guide

How to read PDF File in SAS EG

v9dduhan — Tue, 08 Nov 2016 06:47:16 GMT

Suppose i have call details data in PDF , how can i read it through SAS EG to analyse it further? Any suggession on this ?

I am using SAS EG 5.1 version.

Thanks in Advance

Re: How to read PDF File in SAS EG

Reeza — Tue, 08 Nov 2016 06:49:39 GMT

Do you have SAS Text Miner? That's the only product I know that supports reading a PDF file.

Re: How to read PDF File in SAS EG

Kurt_Bremser — Tue, 08 Nov 2016 06:58:10 GMT

If someone wants you to do analysis on some data, they better supply that data in a suitable format.

Even EG 7.11 does not have the option of importing PDFs.

You might be able to copy/paste the list data out of the PDF into a text file that can be imported.

Re: How to read PDF File in SAS EG

RW9 — Tue, 08 Nov 2016 09:46:45 GMT

PDF is not a data transfer format, it is a rendered picture of some coding information. Trust me, you would be better off typing it all in by hand again rather than try to process it. Go back to the source and ask them to provide the underlying data.

Re: How to read PDF File in SAS EG

Reeza — Tue, 08 Nov 2016 11:21:09 GMT

If you have Adobe Pro there's an option to convert the file to text and/or extract table information to a text file. It's a manual step but can be batched if you have multiple files.

Re: How to read PDF File in SAS EG

RW9 — Tue, 08 Nov 2016 11:23:38 GMT

True. The problem is if you have done anything at all special in the file, it makes it next to impossible to read it without a lot of effort. Even staright cump of text can lead to all kinds of strange spacings and such like to handle.

Re: How to read PDF File in SAS EG

Reeza — Tue, 08 Nov 2016 11:31:19 GMT

The pro conversion helps generate a relatively clean file vs copy and paste. Or the extract table feature.

Re: How to read PDF File in SAS EG

ballardw — Tue, 08 Nov 2016 15:43:29 GMT

Also even the Adobe Pro tools only work if the data was actually converted from something text. I have had to do something similar where the 'data' was actually an image format. Resolution was print to TIFF format then "scan" the resulting image file with character recognition software. Then spend a few hours removing odd characters resulting from poor image quality. That process probably only reduced time from re-typing by about 20 percent. If the document had only been one page instead of 30 I would have retyped it from the start.