Four Bitcoin subgraphs for analysis of multi-input multi-output transactions

This dataset contains 3 transaction subgraphs of the Bitcoin network and 1 aggregated wallet address subgraph. The 3 transaction graphs correspond to the following 3 periods of 49-50 days each. Days are counted in the history of Bitcoin: - 6th April to 25 May 2013 (days 1550-1598) - 8 October to 26 November 2014 (days 2100-2148) - 19 July to 6 September 2016 (days 2750-2799) Every row of the file contains an edge of the transaction network, that is a pair of nodes, separated by a comma. The aggregated wallet address subgraph corresponds to the following period: - 8th December 2010 to 27 January 2011 (days 700-749) Aggregated wallet address means that aggregation heuristics have been applied on the set of wallet addresses to identify that several different addresses are actually likely to belong to the same (set) of user(s). The aggregation is done in two steps: first per transaction, then by transitivity (two sets of aggregated addresses that intersect are merged). Every line of the file contains an edge, a pair of nodes. If you are using the Python Library NetworkX, the graph is uploaded as follows: import networkx as nx G =nx.read_edgelist("AggAddrNetwork700_749.csv") G.order() 6286