erisk2018data

Contains the handler module, which parses the data for a given task
E.g.

from erisk2018data.handler import Task, batchify
task_one = Task('path/to/task1')
training_chunks, test_chunks = task_one.get_split('a')
chunk = training_chunks[0]
length_of_chunk = len(chunk) 
chunk_as_list = list(chunk)
chunk_in_batches = batchify(chunk, batch_size=64)

Install from GitHub

pip install git+https://github.com/BigMiners/eRisk2018/eRisk-2018-parser.git

Building the package

source myvirtualenv/bin/activate
cd navigate/to/erisk-2018-parser
pip install .

Classes

Task

Static methods

__init__(self, path, reader=None)  
    Initialize self.  See help(type(self)) for accurate signature.
    
text_only_xml(path)
    Retrieves the text (titles and content) from a user's writings at a given chunk. Pass this function
    as a parameter to __init__
    :param path: path to xml file
    :return: concatenated titles (from empty posts) and contents

verbatim_xml(path)
    Retrieves the xml from a user's writings at a given chunk. Pass this function
    as a parameter to __init__
    :param path: path to xml file
    :return: xml contents as a string

with_dates_xml(path)
    Retrieves the text (titles and content) and dates from a user's writings at a given chunk. Pass this function
    as a parameter to __init__
    :param path: path to xml file
    :return: concatenated dates, titles (from empty posts) and contents

Instance methods

get_split(self, name)
    Gives a list of 10 generators (1 per chunk) for a given split
    :param name: name of the split
    :return: list of train chunks, list of test chunks