Add Numeric Table Detection samples #48
Conversation
python/14-pdf-numeric-table.ipynb
Outdated
| "file=\"./DemoTable.pdf\"\n", | ||
| "file_key=f\"idp/textract/demo/{os.path.basename(file)}\"\n", | ||
| "s3url=\"s3://\"+data_bucket+\"/\"+file_key\n", | ||
| "!aws s3 cp {file} {s3url} --only-show-errors" |
There was a problem hiding this comment.
I put the sample at: s3://amazon-textract-public-content/code-samples/DemoTable.pdf as public readable. We can get rid of the entire uploading to S3 and Sagemaker code.
Simplification.
There was a problem hiding this comment.
Perfect. I also added a local copy of he file so that the user can see it in the notebook.
| "# Detect Tables in raw text\n", | ||
| "\n", | ||
| "In this example using a threashold of 4 characters out of the 15 most used characters being digits works well to identify if the page is mostly constituted of numeric tables.\n", | ||
| "\n" |
There was a problem hiding this comment.
explain in more detail how this implementation works. This is the key point of the notebook/sample, so we want to be very clear about this. Provide a second implementation as well, so the concept is understood.
There was a problem hiding this comment.
Added a second implementation focusing on each line.
| ] | ||
| }, | ||
| { | ||
| "cell_type": "code", |
There was a problem hiding this comment.
add visualization for the newly created temp.pdf to show that it only contains tables
Issue #, if available:
Description of changes:
I added a static numeric table detection sample in a Jupiter notebook and corrected a missing link in the README
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.