extract_affiliations.py
f3a3b0f515a6master

extract_affiliations.py
View Options

			""" Extracts a list of all affiliations in all the cops (without 7 and 8). Is then used to extract cops 7 and 8.
			Could be refactored to taking the information also from sb meetings or optimized by eliminating errors"""

			import pandas as pd

			# black list coming from frequent errors
			# cancelled out : "national institute","institute for"
			no_affiliation_start = ("ministry ", "bureau ", "executive director", "chairman", "ministerio ", "netherlands committee for", "prof.")
			no_affiliation = ("technology", "protection", "university of vienna")

			affiliations = set([])

			for i in range(1, 26):
			if i != 7 and i != 8:
			data = pd.read_csv("../results/participants-csv/participants_cop" + str(i) + ".csv", encoding="utf-8-sig")
			byaffiliation = data.groupby('affiliation')
			for aff, rest in byaffiliation:
			aff = aff.lower()
			if not aff.startswith(no_affiliation_start) and aff not in no_affiliation:
			affiliations.add(aff)
			# add temporarily
			else:
			print(aff + " --- wrongly extracted from cop" + str(i))

			# write them to a csv file
			df = pd.DataFrame(affiliations)
			df.to_csv('../data/dictionaries/affiliations.csv', encoding="utf-8-sig")