fuzzy-fund-search

Fuzzy Name Search for Hedge Funds

Overview

Fund names in the COVERPAGE.tsv may not match exactly what you're searching for. Fuzzy matching helps find the best match when you have approximate or partial names.

Libraries

Using fuzzywuzzy

pip install fuzzywuzzy python-Levenshtein

Using difflib (built-in)

No installation needed - Python standard library

Search Methods

Method 1: Using fuzzywuzzy (Recommended)

from fuzzywuzzy import fuzz
from fuzzywuzzy import process
import pandas as pd

def fuzzy_search_fund(search_term, coverpage_df, threshold=70):
    """
    Find funds matching a search term using fuzzy matching

    Args:
        search_term: Partial or approximate fund name (e.g., "renaissance technologies")
        coverpage_df: DataFrame from COVERPAGE.tsv
        threshold: Minimum match score (0-100)

    Returns:
        DataFrame of matches sorted by score (highest first)
    """

    fund_names = coverpage_df['FILINGMANAGER_NAME'].tolist()

    # Use extractBests to get multiple matches
    matches = process.extractBests(
        search_term,
        fund_names,
        scorer=fuzz.token_sort_ratio,  # Good for finding partial matches
        score_cutoff=threshold
    )

    # Build result dataframe
    results = []
    for matched_name, score in matches:
        fund_row = coverpage_df[coverpage_df['FILINGMANAGER_NAME'] == matched_name].iloc[0]
        results.append({
            'FILINGMANAGER_NAME': matched_name,
            'ACCESSION_NUMBER': fund_row['ACCESSION_NUMBER'],
            'match_score': score,
            'REPORTCALENDARORQUARTER': fund_row['REPORTCALENDARORQUARTER']
        })

    return pd.DataFrame(results).sort_values('match_score', ascending=False)

# Usage Example
q3_coverpage = pd.read_csv('/root/2025-q3/COVERPAGE.tsv', sep='\t')
matches = fuzzy_search_fund("renaissance technologies", q3_coverpage, threshold=70)
print(matches.head())

# Get best match
best_match = matches.iloc[0]
accession_number = best_match['ACCESSION_NUMBER']
fund_name = best_match['FILINGMANAGER_NAME']