noncesense-research-lab / tx_in_out_distribution Goto Github PK

Using the data posted here originally (see below), assuming number of inputs and number of outputs per transaction are IID random variables drawn from a Negative Binomial distribution, the method of moments for parameter estimation reveals the following.

1. Number of inputs is NB(r,p) with r = 71.900 and p = 0.070, approximately.
2. Number of outputs is NB(r,p) with r = 38.697 and p = 0.137, approximately.

However RingCT presumably has impacted these numbers significantly; the values may be different if we divide the data set into two pieces: pre-RCT and post-RCT.

Would be nice to have an up-to-date data set that excludes pre-RingCT outputs.