filter_by_token_sniffer_score#

API documentation for tradingstrategy.utils.token_filter.filter_by_token_sniffer_score Python function.

filter_by_token_sniffer_score(pairs_df, risk_score, drop_tokens_with_missing_data=True, known_good_tokens={'AAVE', 'DAI', 'FLOKI', 'MKR', 'NEXO', 'PEPE', 'SNX', 'SYN', 'USDC', 'USDS', 'USDT', 'WARB', 'WAVAX', 'WBNB', 'WBTC', 'WETH', 'WMATIC', 'cbBTC'}, max_buy_tax=0.03, max_sell_tax=0.03, printer=<function <lambda>>, taxless_exchanges={'uniswap-v3'})[source]#

Filter out tokens by their TokenSniffer risk score.

Example:

# Load metadata
pairs_df = load_token_metadata(pairs_df, client)

# Scam filter using TokenSniffer
pairs_df = filter_by_token_sniffer_score(pairs_df, 25)
Parameters:

  • known_good_tokens –

    These are set of tokens we know are good, but TokenSniffer gives them bad score.

    E.g. some Coinbase tokens on Base.

  • max_buy_tax –

    Drop tokens with too high buy tax.

    Currently missing tax entries are passed through filter.

  • max_sell_tax –

    Drop tokens with too high sell tax

    Currently missing tax entries are passed through filter.

  • printer – Diagnostics output callback, logger.info() or print.

  • taxless_exchanges – Exchange ids that cannot carry taxed tokens.

  • pairs_df (DataFrame) –

  • risk_score (int) –

Returns:

Pairs DataFrame with too risky pairs removed

Return type:

DataFrame