Hackathon: Gitcoin Open Data Science Hackathon : Sybil Slayer

The data created for the analysis is available here on Hugging Face

The report is in the github repo A time series analysis of gitcoin grants contributors to Grant 15

To run the jupyter notebook please put the features datasets and the hackathon-contributions-dataset_v2.csv as shown in the directory and file structure below:

data
  +----features
          +----- eth_polygon
                    +----- features_eth_polygon.csv    
          +----- eth_std
                    +----- features_eth_std.csv      
  +----- hackathon-contributions-dataset_v2.csv

Use requirements to install the libraries: pip install requirements.txt

The main jupyter notebook used to create the report is k_mean_eth. Be careful before running it is memory heavy because it automatically exports more than 150 charts.

The k_mean_polygon notebook is similar and produces the results for the polygon transactions.

The agglomerative_eth file shows the results in the case of the agglomerative clustering.

An example of clustering for Ethereum is provided in eth_std_clusters.csv, please mind that it was added after the report writings and thus the seed used for K-means was not the same, thus cluster names are different from the one used in the report.

Re-creating the data sets

For transactions

For Etherscan run src\main\extract_etherscan_txs.py

For Polygon chain run src\main\extract_polygon_txs.py

Parameters are hard coded and should be modified!

For extracting the features

run src\main\features\create_feature_df.py Parameters are hard coded and should be modified! You can change the tx_chain parameter to select the network.