score_calculation.py

def score_calculation(weights, school, district):
'''
district is a list of census blocks which we are including on the district
school is a specific school id
the weights normalize the scores taking into account their relative importance
'''
# population of district and minority population from global DataFrame block_data

	pop = block_data['HS_POP'].loc[district].sum()
	minority_pop = block_data['MINORITY'].loc[district].sum()
	#squared error between overall demographic ratio and local demographic ratio
	demographic_score = (abs((minority_pop/pop) - target_ratio))**2*pop*100
	# global distance_matrix holds travel times between each pair of blocks and schools
	# this sum gives the total travel time
	travel_score = sum[distance_matrix[x][school] *
	block_data['HS_POP'].loc[x] for x in district]
	# use graph library NetworkX to find the number of census blocks which the district borders
	outgoing_edges = list(nx.edge_boundary(G,district))
	# the compactnes score is the ratio between the number of bordering districts and its overall size
	compactness_score= len(outgoing_edges)/len(district)
	#weighted sum of scores
	aggregate_score = weights[demographic]*demographic_score
	+ weights[compactness]*compactness_score + weights[travel]*travel_score
	return (demographic_score, travel_score, compactness_score, aggregate_score)