Notes on the files:
===================

citywords_metro_word_cnt_unrestricted_url_filtered_metro_ordered.rpt.gz
	comma-separated values in GZIP compression
	column headers:
		word,cbsacode,population,cnt

aggregate_metrics_data_w_fit_results.json
	each line is a JSON containing the data and results of the power-law fits for the aggregate metrics
	data is contained in keys "x" (population) and "y" (metrics)
	"model" stands for the model from Leitao et al. 2016

wordcount_fits_results.json.gz
	gzipped line by line JSON file
	contains data and results for the power-law fits of words
	data is contained in keys "x" (population) and "y" (metrics)
	Delta BIC is calculated as compared to the total word volume, e.g. a fixed gamma model from Leitao et al. with 1.0207

zipf.csv
	comma-separated values
	contains how many times a certain word count occurred in the total corpus
	column headers:
		total_word_count,count_of_total_word_count

citywise_zipf.csv
	comma-separated values
	contains the results of Zipf's law fits on citywise corpora
	column headers:
		cbsacode,beta,beta_error,total_wordcount