Skip to content

Documentation for comparison_helpers functions

The comparison_helpers functions are a set of functions to help users create better comparisons by helping them understand string comparators (fuzzy matching) and phonetic matching.

The detailed API for each of these are outlined below.

Comparison Helpers API

comparator_score(str1, str2, decimal_places=2)

Helper function to give the similarity between two strings for the string comparators in splink.

Examples:

import splink.comparison_helpers as ch

ch.comparator_score("Richard", "iRchard")
Source code in splink/comparison_helpers.py
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
def comparator_score(str1, str2, decimal_places=2):
    """Helper function to give the similarity between two strings for
    the string comparators in splink.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        ch.comparator_score("Richard", "iRchard")
        ```
    """
    con = duckdb.connect()

    sql = f"""
        select
        '{str1}' as string1,
        '{str2}' as string2,
        {comparator_cols_sql.format(
            comparison1 = 'string1',
            comparison2 = 'string2',
            decimal_places=decimal_places
        )}
    """
    return con.execute(sql).fetch_df()

comparator_score_chart(list, col1, col2)

Helper function returning a heatmap showing the sting similarity scores and string distances for a list of strings.

Examples:

import splink.comparison_helpers as ch

list = {
        "string1": ["Stephen", "Stephen", "Stephen"],
        "string2": ["Stephen", "Steven", "Stephan"],
        }

ch.comparator_score_chart(list, "string1", "string2")
Source code in splink/comparison_helpers.py
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
def comparator_score_chart(list, col1, col2):
    """Helper function returning a heatmap showing the sting similarity
    scores and string distances for a list of strings.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        list = {
                "string1": ["Stephen", "Stephen", "Stephen"],
                "string2": ["Stephen", "Steven", "Stephan"],
                }

        ch.comparator_score_chart(list, "string1", "string2")
        ```
    """

    df = comparator_score_df(list, col1, col2)

    df["strings_to_compare"] = df["string1"] + ", " + df["string2"]

    df_long = pd.melt(
        df,
        id_vars=["strings_to_compare"],
        value_vars=[
            "jaro_similarity",
            "jaro_winkler_similarity",
            "jaccard_similarity",
            "levenshtein_distance",
            "damerau_levenshtein_distance",
        ],
        var_name="comparator",
        value_name="score",
    )

    similarity_df = df_long[df_long["comparator"].str.contains("similarity")]
    similarity_df.loc[:, "comparator"] = similarity_df["comparator"].str.replace(
        "_similarity", ""
    )
    similarity_records = similarity_df.to_json(orient="records")
    distance_df = df_long[df_long["comparator"].str.contains("distance")]
    distance_df.loc[:, "comparator"] = distance_df["comparator"].str.replace(
        "_distance", ""
    )
    distance_records = distance_df.to_json(orient="records")

    return _comparator_score_chart(similarity_records, distance_records)

comparator_score_df(list, col1, col2, decimal_places=2)

Helper function returning a dataframe showing the string similarity scores and string distances for a list of strings.

Examples:

import splink.comparison_helpers as ch

list = {
        "string1": ["Stephen", "Stephen","Stephen"],
        "string2": ["Stephen", "Steven", "Stephan"],
        }

ch.comparator_score_df(list, "string1", "string2")
Source code in splink/comparison_helpers.py
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
def comparator_score_df(list, col1, col2, decimal_places=2):
    """Helper function returning a dataframe showing the string similarity
    scores and string distances for a list of strings.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        list = {
                "string1": ["Stephen", "Stephen","Stephen"],
                "string2": ["Stephen", "Steven", "Stephan"],
                }

        ch.comparator_score_df(list, "string1", "string2")
        ```
    """
    duckdb.connect()

    list = pd.DataFrame(list)

    sql = f"""
        select
        {col1}, {col2},
        {comparator_cols_sql.format(
            comparison1 = col1,
            comparison2 = col2,
            decimal_places=decimal_places
        )},
        from list
    """

    return duckdb.sql(sql).df()

comparator_score_threshold_chart(list, col1, col2, similarity_threshold=None, distance_threshold=None)

Helper function returning a heatmap showing the sting similarity scores and string distances for a list of strings given a threshold.

Examples:

import splink.comparison_helpers as ch

list = {
        "string1": ["Stephen", "Stephen","Stephen"],
        "string2": ["Stephen", "Steven", "Stephan"],
        }

ch.comparator_score_threshold_chart(data,
                         "string1", "string2",
                         similarity_threshold=0.8,
                         distance_threshold=2)
Source code in splink/comparison_helpers.py
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
def comparator_score_threshold_chart(
    list, col1, col2, similarity_threshold=None, distance_threshold=None
):
    """Helper function returning a heatmap showing the sting similarity
    scores and string distances for a list of strings given a threshold.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        list = {
                "string1": ["Stephen", "Stephen","Stephen"],
                "string2": ["Stephen", "Steven", "Stephan"],
                }

        ch.comparator_score_threshold_chart(data,
                                 "string1", "string2",
                                 similarity_threshold=0.8,
                                 distance_threshold=2)
        ```
    """
    df = comparator_score_df(list, col1, col2)

    df["strings_to_compare"] = df["string1"] + ", " + df["string2"]

    df_long = pd.melt(
        df,
        id_vars=["strings_to_compare"],
        value_vars=[
            "jaro_similarity",
            "jaro_winkler_similarity",
            "jaccard_similarity",
            "levenshtein_distance",
            "damerau_levenshtein_distance",
        ],
        var_name="comparator",
        value_name="score",
    )

    similarity_df = df_long.loc[df_long["comparator"].str.contains("similarity"), :]
    similarity_df["comparator"] = similarity_df["comparator"].str.replace(
        "_similarity", ""
    )
    similarity_records = similarity_df.to_json(orient="records")
    distance_df = df_long.loc[df_long["comparator"].str.contains("distance"), :]
    distance_df["comparator"] = distance_df["comparator"].str.replace("_distance", "")
    distance_records = distance_df.to_json(orient="records")

    return _comparator_score_threshold_chart(
        similarity_records, distance_records, similarity_threshold, distance_threshold
    )

phonetic_match_chart(list, col1, col2)

Helper function returning a heatmap showing the phonetic transform and matches for a list of strings given a threshold.

Examples:

import splink.comparison_helpers as ch

list = {
        "string1": ["Stephen", "Stephen","Stephen"],
        "string2": ["Stephen", "Steven", "Stephan"],
        }

ch.comparator_score_threshold_chart(list,
                         "string1", "string2",
                         similarity_threshold=0.8,
                         distance_threshold=2)
Source code in splink/comparison_helpers.py
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
def phonetic_match_chart(list, col1, col2):
    """Helper function returning a heatmap showing the phonetic transform and
    matches for a list of strings given a threshold.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        list = {
                "string1": ["Stephen", "Stephen","Stephen"],
                "string2": ["Stephen", "Steven", "Stephan"],
                }

        ch.comparator_score_threshold_chart(list,
                                 "string1", "string2",
                                 similarity_threshold=0.8,
                                 distance_threshold=2)
        ```
    """

    df = phonetic_transform_df(list, "string1", "string2")

    df["strings_to_compare"] = df["string1"] + ", " + df["string2"]

    df_long = pd.melt(
        df,
        id_vars=["strings_to_compare"],
        value_vars=[
            "metaphone",
            "dmetaphone",
            "soundex",
        ],
        var_name="phonetic",
        value_name="transform",
    )
    df_long["match"] = df_long["transform"].apply(lambda x: x[0] == x[1])

    records = df_long.to_json(orient="records")

    return _phonetic_match_chart(records)

phonetic_transform(string)

Helper function to give the phonetic transformation of two strings with Soundex, Metaphone and Double Metaphone.

Examples:

phonetic_transform("Richard", "iRchard")
Source code in splink/comparison_helpers.py
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
def phonetic_transform(string):
    """Helper function to give the phonetic transformation of two strings with
    Soundex, Metaphone and Double Metaphone.

    Examples:
        ```py
        phonetic_transform("Richard", "iRchard")
        ```
    """
    transforms = {}

    # Soundex Transform
    soundex_transform = phonetics.soundex(string)
    transforms["soundex"] = soundex_transform

    # Metaphone distance
    metaphone_transform = phonetics.metaphone(string)
    transforms["metaphone"] = metaphone_transform

    # Metaphone distance
    dmetaphone_transform = phonetics.dmetaphone(string)
    transforms["dmetaphone"] = dmetaphone_transform

    return transforms

phonetic_transform_df(list, col1, col2)

Helper function returning a dataframe showing the phonetic transforms for a list of strings.

Examples:

import splink.comparison_helpers as ch

list = {
        "string1": ["Stephen", "Stephen","Stephen"],
        "string2": ["Stephen", "Steven", "Stephan"],
        }

ch.phonetic_match_chart(list, "string1", "string2")
Source code in splink/comparison_helpers.py
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
def phonetic_transform_df(list, col1, col2):
    """Helper function returning a dataframe showing the phonetic transforms
    for a list of strings.

    Examples:
        ```py
        import splink.comparison_helpers as ch

        list = {
                "string1": ["Stephen", "Stephen","Stephen"],
                "string2": ["Stephen", "Steven", "Stephan"],
                }

        ch.phonetic_match_chart(list, "string1", "string2")
        ```
    """

    df = pd.DataFrame(list)

    df[f"soundex_{col1}"] = df.apply(lambda row: phonetics.soundex(row[col1]), axis=1)
    df[f"soundex_{col2}"] = df.apply(lambda row: phonetics.soundex(row[col2]), axis=1)
    df[f"metaphone_{col1}"] = df.apply(
        lambda row: phonetics.metaphone(row[col1]), axis=1
    )
    df[f"metaphone_{col2}"] = df.apply(
        lambda row: phonetics.metaphone(row[col2]), axis=1
    )
    df[f"dmetaphone_{col1}"] = df.apply(
        lambda row: phonetics.dmetaphone(row[col1]), axis=1
    )
    df[f"dmetaphone_{col2}"] = df.apply(
        lambda row: phonetics.dmetaphone(row[col2]), axis=1
    )

    df["soundex"] = df.apply(
        lambda x: [x[f"soundex_{col1}"], x[f"soundex_{col2}"]], axis=1
    )
    df["metaphone"] = df.apply(
        lambda x: [x[f"metaphone_{col1}"], x[f"metaphone_{col2}"]], axis=1
    )
    df["dmetaphone"] = df.apply(
        lambda x: [x[f"dmetaphone_{col1}"], x[f"dmetaphone_{col2}"]], axis=1
    )

    phonetic_df = df[[col1, col2, "soundex", "metaphone", "dmetaphone"]]

    return phonetic_df