Addresses that claimed during early decay periods received a smaller claimable amount than expected.

To investigate this issue, all of the claims and merge_claim_records events that ocurred in history where collected and then analyzed in this repo through pandas and jupyter notebook.

Details about how the data was collected can be found at decay-data.

The end goal was to calculate the amount that each sender of the triggered claim events in history had failed to receive due to a decay bug. In other words:

how much should the user had received - how much did the user actually received

To achieve this the following steps were followed:

1. Collect on a database all the claim and merge_claim_records events that occurred in Evmos’s history using https://github.com/facs95/decay-data.
   • This process is explained in detail within the mentioned repo.
2. Data was cleaned and filtered using pandas to generate a table with the following columns:
   • sender - the address that received the claim rewards
   • height - the height at which the event occurred
   • event_type - action type of the event that got triggered
     • ACTION_EVM
     • ACTION_DELEGATE
     • ACTION_VOTE
     • ACTION_IBC_TRANSFER
   • initial_claimable_amount - claim amount available at genesis for the sender if any
   • claimable_at_event - the amount available to claim for the sender at the event if any.
     • If there were any merge_claims_records events present for the sender account before the event height, then the claimable amount at genesis of both accounts were taking into account and added.
   • amount_claimed - the amount the address actually received on the event
   • lost_amount - the difference between claimable_at_event and amount_claimed
     • This is the amount that should be returned.

Details about how this information was processed can be found under ./decay-data.ipynb

• Jupyter
• Pandas
• Numpy
• Sqlite3

The results are exposed under decay_lost_results.csv table.