Hungarian ANUM Numbers

Introduction

The function clean_hu_anum() cleans a column containing Hungarian ANUM number (ANUM) strings, and standardizes them in a given format. The function validate_hu_anum() validates either a single ANUM strings, a column of ANUM strings or a DataFrame of ANUM strings, returning True if the value is valid, and False otherwise.

ANUM strings can be converted to the following formats via the output_format parameter:

• compact: only number strings without any seperators or whitespace, like “12892312”

• standard: ANUM strings with proper whitespace in the proper places. Note that in the case of ANUM, the compact format is the same as the standard one.

Invalid parsing is handled with the errors parameter:

• coerce (default): invalid parsing will be set to NaN

• ignore: invalid parsing will return the input

• raise: invalid parsing will raise an exception

The following sections demonstrate the functionality of clean_hu_anum() and validate_hu_anum().

An example dataset containing ANUM strings

import pandas as pd
import numpy as np
df = pd.DataFrame(
    {
        "anum": [
            'HU-12892312',
            'HU-12892313',
            'BE 428759497',
            'BE431150351',
            "002 724 334",
            "hello",
            np.nan,
            "NULL",
        ],
        "address": [
            "123 Pine Ave.",
            "main st",
            "1234 west main heights 57033",
            "apt 1 789 s maple rd manhattan",
            "robie house, 789 north main street",
            "1111 S Figueroa St, Los Angeles, CA 90015",
            "(staples center) 1111 S Figueroa St, Los Angeles",
            "hello",
        ]
    }
)
df

	anum	address
0	HU-12892312	123 Pine Ave.
1	HU-12892313	main st
2	BE 428759497	1234 west main heights 57033
3	BE431150351	apt 1 789 s maple rd manhattan
4	002 724 334	robie house, 789 north main street
5	hello	1111 S Figueroa St, Los Angeles, CA 90015
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles
7	NULL	hello

1. Default clean_hu_anum

By default, clean_hu_anum will clean anum strings and output them in the standard format with proper separators.

from dataprep.clean import clean_hu_anum
clean_hu_anum(df, column = "anum")

	anum	address	anum_clean
0	HU-12892312	123 Pine Ave.	12892312
1	HU-12892313	main st	NaN
2	BE 428759497	1234 west main heights 57033	NaN
3	BE431150351	apt 1 789 s maple rd manhattan	NaN
4	002 724 334	robie house, 789 north main street	NaN
5	hello	1111 S Figueroa St, Los Angeles, CA 90015	NaN
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles	NaN
7	NULL	hello	NaN

2. Output formats

This section demonstrates the output parameter.

standard (default)

clean_hu_anum(df, column = "anum", output_format="standard")

	anum	address	anum_clean
0	HU-12892312	123 Pine Ave.	12892312
1	HU-12892313	main st	NaN
2	BE 428759497	1234 west main heights 57033	NaN
3	BE431150351	apt 1 789 s maple rd manhattan	NaN
4	002 724 334	robie house, 789 north main street	NaN
5	hello	1111 S Figueroa St, Los Angeles, CA 90015	NaN
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles	NaN
7	NULL	hello	NaN

compact

clean_hu_anum(df, column = "anum", output_format="compact")

	anum	address	anum_clean
0	HU-12892312	123 Pine Ave.	12892312
1	HU-12892313	main st	NaN
2	BE 428759497	1234 west main heights 57033	NaN
3	BE431150351	apt 1 789 s maple rd manhattan	NaN
4	002 724 334	robie house, 789 north main street	NaN
5	hello	1111 S Figueroa St, Los Angeles, CA 90015	NaN
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles	NaN
7	NULL	hello	NaN

3. inplace parameter

This deletes the given column from the returned DataFrame. A new column containing cleaned ANUM strings is added with a title in the format "{original title}_clean".

clean_hu_anum(df, column="anum", inplace=True)

	anum_clean	address
0	12892312	123 Pine Ave.
1	NaN	main st
2	NaN	1234 west main heights 57033
3	NaN	apt 1 789 s maple rd manhattan
4	NaN	robie house, 789 north main street
5	NaN	1111 S Figueroa St, Los Angeles, CA 90015
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles
7	NaN	hello

4. errors parameter

coerce (default)

clean_hu_anum(df, "anum", errors="coerce")

	anum	address	anum_clean
0	HU-12892312	123 Pine Ave.	12892312
1	HU-12892313	main st	NaN
2	BE 428759497	1234 west main heights 57033	NaN
3	BE431150351	apt 1 789 s maple rd manhattan	NaN
4	002 724 334	robie house, 789 north main street	NaN
5	hello	1111 S Figueroa St, Los Angeles, CA 90015	NaN
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles	NaN
7	NULL	hello	NaN

ignore

clean_hu_anum(df, "anum", errors="ignore")

	anum	address	anum_clean
0	HU-12892312	123 Pine Ave.	12892312
1	HU-12892313	main st	HU-12892313
2	BE 428759497	1234 west main heights 57033	BE 428759497
3	BE431150351	apt 1 789 s maple rd manhattan	BE431150351
4	002 724 334	robie house, 789 north main street	002 724 334
5	hello	1111 S Figueroa St, Los Angeles, CA 90015	hello
6	NaN	(staples center) 1111 S Figueroa St, Los Angeles	NaN
7	NULL	hello	NaN

4. validate_hu_anum()

validate_hu_anum() returns True when the input is a valid ANUM. Otherwise it returns False.

The input of validate_hu_anum() can be a string, a Pandas DataSeries, a Dask DataSeries, a Pandas DataFrame and a dask DataFrame.

When the input is a string, a Pandas DataSeries or a Dask DataSeries, user doesn’t need to specify a column name to be validated.

When the input is a Pandas DataFrame or a dask DataFrame, user can both specify or not specify a column name to be validated. If user specify the column name, validate_hu_anum() only returns the validation result for the specified column. If user doesn’t specify the column name, validate_hu_anum() returns the validation result for the whole DataFrame.

from dataprep.clean import validate_hu_anum
print(validate_hu_anum("HU-12892312"))
print(validate_hu_anum("HU-12892313"))
print(validate_hu_anum('BE 428759497'))
print(validate_hu_anum('BE431150351'))
print(validate_hu_anum("004085616"))
print(validate_hu_anum("hello"))
print(validate_hu_anum(np.nan))
print(validate_hu_anum("NULL"))

True
False
False
False
False
False
False
False

Series

validate_hu_anum(df["anum"])

0     True
1    False
2    False
3    False
4    False
5    False
6    False
7    False
Name: anum, dtype: bool

DataFrame + Specify Column

validate_hu_anum(df, column="anum")

0     True
1    False
2    False
3    False
4    False
5    False
6    False
7    False
Name: anum, dtype: bool

Only DataFrame

validate_hu_anum(df)

	anum	address
0	True	False
1	False	False
2	False	False
3	False	False
4	False	False
5	False	False
6	False	False
7	False	False