pandas-fuzz 0.1.3

Extension for `pandas` to use `rapidfuzz` for fuzzy matching.

pandas-fuzz

---

Extension for pandas to use rapidfuzz for fuzzy matching.

Requirements

• Python 3.9 or later
• On Windows the Visual C++ 2019 redistributable is required

Installation

pip install pandas_fuzz

Usage

To register the extension make sure to import pandas_fuzz before using it`.

import pandas as pd
import pandas_fuzz

Alternatively, you can import pandas from pandas_fuzz directly.

from pandas_fuzz import pandas as pd

rapidfuzz.fuzz

pandas_fuzz integrates the following functions from rapidfuzz.fuzz into pandas. These functions are available in the fuzz namespace for both pandas.Series and pandas.DataFrame.

• rapidfuzz.fuzz.ratio
• rapidfuzz.fuzz.partial_ratio
• rapidfuzz.fuzz.partial_ratio_alignment
• rapidfuzz.fuzz.token_sort_ratio
• rapidfuzz.fuzz.token_set_ratio
• rapidfuzz.fuzz.token_ratio
• rapidfuzz.fuzz.partial_token_sort_ratio
• rapidfuzz.fuzz.partial_token_set_ratio
• rapidfuzz.fuzz.partial_token_ratio
• rapidfuzz.fuzz.WRatio
• rapidfuzz.fuzz.QRatio

pandas.Series

apply fuzz.ratio element wise to pd.Series.

>>> pd.Series(["this is a test", "this is a test!"]).fuzz.ratio("this is a test!")
0     96.551724
1    100.000000
dtype: float64

pandas.DataFrame

apply fuzz.ratio row wise to columns s1 and s2

>>> pd.DataFrame({
    "s1": ["this is a test", "this is a test!"],
    "s2": ["this is a test", "this is a test!"]
}).fuzz.ratio("s1", "s2")
0    100.0
1    100.0
dtype: float64

Dependencies

---