fix(clean): add comma after street suffix or name

sfu-db · Mar 12, 2021 · e7655db · e7655db
1 parent 1c03d8b
commit e7655db
Show file tree

Hide file tree

Showing 3 changed files with 18 additions and 13 deletions.
diff --git a/dataprep/clean/clean_address.py b/dataprep/clean/clean_address.py
@@ -29,7 +29,7 @@ def clean_address(
     df: Union[pd.DataFrame, dd.DataFrame],
     column: str,
     output_format: str = "(building) house_number street_prefix_abbr "
-    "street_name street_suffix_abbr apartment, city, state_abbr zipcode",
+    "street_name street_suffix_abbr, apartment, city, state_abbr zipcode",
     must_contain: Tuple[str, ...] = ("house_number", "street_name"),
     split: bool = False,
     inplace: bool = False,
@@ -66,7 +66,7 @@ def clean_address(
         The output_format can contain '\\\\t' characters to specify how to split the output into
         columns.
 
-        (default: '(building) house_number street_prefix_abbr street_name street_suffix_abbr
+        (default: '(building) house_number street_prefix_abbr street_name street_suffix_abbr,
         apartment, city, state_abbr zipcode')
     must_contain
         A tuple containing parts of the address that must be included for the address to be
@@ -299,6 +299,11 @@ def _address_dict_to_string(address: Dict[str, str], output_format: str, split:
     if split:
         output_format = "\t".join(output_format.split())
 
+    # add a comma after the street name if there is no street suffix
+    # in address_items
+    if "street_suffix_abbr" not in address_items and not split:
+        output_format = output_format.replace("street_name", "street_name,")
+
     # first split output_format into each column of the final output
     # for each column split it into attributes and add the corresponding
     # cleaned part of the address to the output for each attribute

diff --git a/dataprep/tests/clean/test_clean_address.py b/dataprep/tests/clean/test_clean_address.py
@@ -38,11 +38,11 @@ def test_clean_default(df_addresses: pd.DataFrame) -> None:
     df_check = df_addresses.copy()
     df_check["messy_address_clean"] = [
         "123 Pine Ave.",
-        "1234 W. Main Hts. 57033",
+        "1234 W. Main Hts., 57033",
         np.nan,
-        "(Robie House) 789 N. Main St. Manhattan, NY",
-        "1111 S. Figueroa St. Los Angeles, CA 90015",
-        "(Staples Center) 1111 S. Figueroa St. Los Angeles",
+        "(Robie House) 789 N. Main St., Manhattan, NY",
+        "1111 S. Figueroa St., Los Angeles, CA 90015",
+        "(Staples Center) 1111 S. Figueroa St., Los Angeles",
         np.nan,
         np.nan,
         np.nan,
@@ -133,11 +133,11 @@ def test_clean_must_contain(df_addresses: pd.DataFrame) -> None:
     df_check["messy_address_clean"] = [
         np.nan,
         np.nan,
-        "S. Maple Rd. Apt 1, Manhattan",
-        "(Robie House) 789 N. Main St. Manhattan, NY",
-        "1111 S. Figueroa St. Los Angeles, CA 90015",
-        "(Staples Center) 1111 S. Figueroa St. Los Angeles",
-        "S. Figueroa Los Angeles",
+        "S. Maple Rd., Apt 1, Manhattan",
+        "(Robie House) 789 N. Main St., Manhattan, NY",
+        "1111 S. Figueroa St., Los Angeles, CA 90015",
+        "(Staples Center) 1111 S. Figueroa St., Los Angeles",
+        "S. Figueroa, Los Angeles",
         np.nan,
         np.nan,
         np.nan,

diff --git a/docs/source/user_guide/clean/clean_address.ipynb b/docs/source/user_guide/clean/clean_address.ipynb
@@ -43,7 +43,7 @@
     "* state_full: ('California')\n",
     "* zipcode: ('57903')\n",
     "\n",
-    "The default `output_format` is \"(building) house_number street_prefix_abbr street_name street_suffix_abbr apartment,\n",
+    "The default `output_format` is \"(building) house_number street_prefix_abbr street_name street_suffix_abbr, apartment,\n",
     "     city, state_abbr zipcode\"\n",
     "     \n",
     "The `must_contain` parameter takes a tuple containing parts of the address that must be included for the address to be successfully cleaned, the following keywords are supported.\n",
@@ -363,7 +363,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.6"
+   "version": "3.9.1"
   }
  },
  "nbformat": 4,