Merge branch 'develop'

ihmeuw · Apr 1, 2023 · 182c4f7 · 182c4f7
2 parents ac7a463 + 7d2741c
commit 182c4f7
Show file tree

Hide file tree

Showing 5 changed files with 12 additions and 6 deletions.
diff --git a/CHANGELOG.rst b/CHANGELOG.rst
@@ -1,3 +1,7 @@
+**0.2.1 - 03/31/23**
+
+ - Fix bug preventing generation of W2/1099 forms
+
 **0.2.0 - 03/31/23**
 
  - Implemented W2/1099 forms

diff --git a/src/pseudopeople/__about__.py b/src/pseudopeople/__about__.py
@@ -13,7 +13,7 @@
 __summary__ = "pseudopeople is package which adds noise to simulated census-scale data using standard scientific Python tools."
 __uri__ = "https://github.com/ihmeuw/pseudopeople"
 
-__version__ = "0.2.0"
+__version__ = "0.2.1"
 
 __author__ = "The pseudopeople developers"
 __email__ = "[email protected]"

diff --git a/src/pseudopeople/noise_functions.py b/src/pseudopeople/noise_functions.py
@@ -59,17 +59,21 @@ def generate_incorrect_selections(
     :returns: pd.Series where data has been noised with other values from a list of possibilities
     """
 
-    col = column.name
+    selection_type = {
+        "employer_state": "state",
+        "mailing_address_state": "state",
+    }.get(str(column.name), column.name)
+
     selection_options = pd.read_csv(paths.INCORRECT_SELECT_NOISE_OPTIONS_DATA)
 
     # Get possible noise values
     # todo: Update with exclusive resampling when vectorized_choice is improved
-    options = selection_options.loc[selection_options[col].notna(), col]
+    options = selection_options.loc[selection_options[selection_type].notna(), selection_type]
     new_values = vectorized_choice(
         options=options,
         n_to_choose=len(column),
         randomness_stream=randomness_stream,
-        additional_key=f"{additional_key}_{col}_incorrect_select_choice",
+        additional_key=f"{additional_key}_{column.name}_incorrect_select_choice",
     ).to_numpy()
 
     return pd.Series(new_values, index=column.index)

diff --git a/tests/unit/test_column_noise.py b/tests/unit/test_column_noise.py
@@ -55,7 +55,6 @@ def default_configuration():
 
 
 def test_generate_missing_data(dummy_dataset):
-
     config = get_configuration()
     config.update(
         {

diff --git a/tests/unit/test_noise_form.py b/tests/unit/test_noise_form.py
@@ -176,7 +176,6 @@ def test_correct_forms_are_used(func, form, mocker):
 
 
 def test_two_noise_functions_are_independent(mocker):
-
     # Make simple config tree to test 2 noise functions work together
     config_tree = ConfigTree(
         {
-Original file line number
+Diff line change
@@ Expand Up / @@ -55,7 +55,6 @@ def default_configuration(): @@
     def test_generate_missing_data(dummy_dataset):
         config = get_configuration()
         config.update(
             {
@@ Expand Down @@