Group training samples by page about edgar-doc-parser HOT 3 CLOSED

@kamilkrukowski
in the TODO，

TODO: Convert to list of [lists of tuples, page number, parent_company_tikr] scheme where each list of tuples consists of all annotated webelements on that page

do we need to specify the document for the lists of tuples?
i.e. [lists of tuples, doc_reference,page number, parent_company_tikr]
doc_reference is the value from the parser.get_annotated_submissions(tikr)

from edgar-doc-parser.

from edgar-doc-parser.

[ #i.e. this is page L of document N of company M, but those 3 variables are hidden within the features of x
(x: value, y: [label_list]),
(x: value, y: [label_list]),
...
]
for now value can refer to webelement.text, later on we can expand the 'features'

The core functionality is really the grouping --> The features will be handled elsewhere later.

from edgar-doc-parser.